f>]p |::nrg|..:riPQ
^l../- ^'. J:. . .h./'il. .:,..J;!,„<fh../
'Smn&d on
t.c> tJae M'enioiry ol
IBIbert llilalvcr A.jb,lstrO'B"i
rbe Sympo^wm Vv^as hdd Awgust IS-IS, 1983
SporKKM'w] by tba
JuitioDatl jfi-iaiine H;slKde!ii Scrvia?
ABiiiCi'itt-Hi', Society «.)';" ]l':;:!i i:hyoloi?j:^ti5 aad HerfK'rtc^loBiBi'j
nj
_D
■=0
3
r^
L
1— 1
D
o
m
□
c, /
QNTOGENY AND SYSTEMATICS
^ OF FISHES
Based on
An International Symposium Dedicated
to the Memory of
Elbert Halvor Ahlstrom
The Symposium was held August 15-18, 1983, La JoUa, CaUfomia
Sponsored by the
National Marine Fisheries Service
National Oceanic and Atmospheric Administration
United States Department of Commerce
\9
■0
4
0
5:^
Special Publication Number 1
American Society of Ichthyologists and Herpetologists
Library of Congress Catalogue Card Number: 84-72702
ISSN No. 0748-0539
© Copyright, 1984, by The American Society of Ichthyologists and Herpetologists
Pnnted by Allen Press Inc.. Lawrence, KS 66044 USA
Preface
The National Marine Fisheries Service organized, supported and conducted an international symposium entitled
Ontogeny and Systematics of Fishes, held in La JoUa, California on August 15-18, 1983, and dedicated to the memory
of Elbert Halvor Ahlstrom. Dr. R, Lasker served as convener. The papers presented at that symposium form the basis
for this book, which is published by the American Society of Ichthyologists and Herpetologists as their Supplement to
Copeia, Special Publication Number 1 . Financial support was provided by the National Marine Fisheries Service, National
Oceanic and Atmospheric Administration, U.S. Department of Commerce.
For many years. Dr. Ahlstrom planned to write a book on larval fishes and ways in which they contributed to systematics.
A few years before his untimely death, he and his colleague H. G. Moser outlined such a book and began to work on
the initial chapters. Dr. Ahlstrom left a vast store of notes, data, and partly completed manuscnpts. Dr. Moser realized
that much of the significance of these unique and important data would be lost unless they were brought to light. He
approached colleagues at the Southwest Fisheries Center to gather a group of larval fish workers who had worked closely
with Dr. Ahlstrom, and who were given access to his notes, to collaborate on the book. From this initiative a plan
developed to conduct a symposium and publish the results in a book to accomplish the original plan of Dr. Ahlstrom
and honor his memory as one of the nation's foremost fishery scientists.
A symposium steering committee was formed with H. G. Moser as Chairman and consisted of D. M. Cohen, M. P.
Fahay, A. W. Kendall, Jr.. W. J. Richards and S. L. Richardson. The steering committee first met in Boulder, Colorado
to develop an outline for the symposium and book and invite potential contributors. The aim was to present the current
state of knowledge of early life history of fishes and apply that to systematics. Originally it was intended to concentrate
solely on the marine groups with which Dr. Ahlstrom had worked, but because of recent advances in freshwater and
other early life history work, the plan was expanded to include all but the primitive osteoglossomorphs. Thus, the coverage
was to start with the elopomorphs.
Following the Boulder meeting, potential contributors were contacted and responded enthusiastically. The Steering
Committee met subsequently in Ocean Springs, Mississippi and Miami, Florida to review progress and refine plans.
Because of the subject matter it seemed appropriate that the American Society of Ichthyologists and Herpetologists
collaborate in publishing the papers resulting from the symposium. C. R. Robins, then President of ASIH, supported
this suggestion and assisted in many ways. Subsequent to the symposium, manuscripts were reviewed and edited by the
Steering Committee of the Symposium, which served as an editorial committee for this volume.
The Steering Committee thanks all of the authors of this volume among whom there was a great exchange of ideas
and generous help. Much additional assistance was provided to the authors and is here acknowledged. Institutional
support was provided by the National Marine Fisheries Service through contributions from each of the four Fisheries
Centers— Southwest, Southeast, Northwest and Alaska and Northeast. Support was provided by the National Science
Foundation through grants DEB76-82279, DEB78-26540; the National Geographic Society by grant 2535-82 from the
Committee for Research and Exploration; the Robert E. Maytag Fellowship at the University of Miami; Natural History
Museum of Los Angeles County; the Australian Museum Trust, the Australian Marine Science and Technologies Advisory
Committee, the Commonwealth Science and Industrial Research Organization Science and Industry Endowment Fund,
and the employers of the contributors.
The following individuals supplied specimens, data, technical assistance, publications, and reviewed drafts of manu-
scripts: M. Allen. R. M. Allen, A. Alvarino, D. Ambrose, M. E. Anderson, W. D. Anderson. Jr.. F. Balbontin. C. Baldwin,
E. K. Balon, P. Berrien, D. Blood, S. Boardman, S. S. Boggs, E. Bohlke, M. Bradbury, J. Brill, D. Brown, J. Bullock, M.
S. Busby. J. A. Cambray, P. Camus, M. H. Carrington, B. Chemoff, T. A. Clarke, M. Culbreth, M. Cluxton, S. Coombs,
A. S. Creighton, K. Davis, W. P. Davis. C. E. Dawson. M. Dehaan, N. Demir, A. Desai, H. H. DeWitt, M. DeWitt, Y.
Dotsu, S. D'Vincent, B. R. Engstrand, D. Faber, N. R. Foster, P. Fourmanoir, C. Frandsen, H. J. Franke, E. Fridgeirsson,
W. George, R. H. Gibbs, G. Gilmore, D. Gittings, W. Gladstone, T. Goh. M. F. Gomon. B. Goldman, A. R. Gosline,
W. A. Gosline, A. E. Gosztonyi, P. H. Greenwood, D. Haggner, G. R. Harbison, G. S. Hardy, K. Hartel. R. Hartwick,
T. Hecht, E. Hubert. J. M. Humphries. J. C. Hureau. T. Iwamoto. S. Jewett, P. Keener, S. Kelley, F. Kirschbaum, N.
Komada, Y. Konishi, D. L. Kramer, J. K. Langhammer. K. Lazara, K. Lee, S. Lincoln, J. Lobon-Cervia, V. J. Loeb, G.
Lundy, N. A. Mackintosh, F. Mago-Leccia, A. M. Martinez, D. McAllister, M. McCabe, J. McCosker. R. F. McGinnis.
R. McMichael. R. Meier. N. Merrett. J. Michalski, J. Mighell, R. R. Miller, C. Mills, A. Miskiewicz, G. E. E. Moodie,
K. H. Moore, K. Mori, J. Moyer, J. A. Musick, T. Nakata, G. Nelson, J. Nelson, J. Nichols, J. Nielsen, T. North, S.
Ochman, G. Patchell, L. R. Parenti, K. Peters. T. Pomeranz, S. Poss, L. C. Prescott, J. Quast, J. Randall, K. S. Raymond,
B. Remington, C. S. Richards, T. Roberts, D. E. Rosen. R. Schoknecht, A. Sekerak, T. Senta, J. Shapiro. J. Shoemaker,
P. L. Shafland, M. Shiogaki, D. L. Schultz, P. H. Skelton, P. E. Smith, J. Song, D. E. Snyder, A. Soeldner, C. Stehr, D.
Stein, B. Stender, K. Steward, K. Stoddard, R. E. Strauss, G. Stroud, K. J. Sulak, A. Suzumoto, H. Sweatman, J. N.
Taylor, V. R. Thomas, G. Theilacker. R. Thresher, R. Triemer, D. Tweedle, J. C. Tyler, F. Utter, F. Van Dolah, R.
Vari, B. Vinter. L. Vlyman, R. Wallus. T. Watanabe. B. A. Watkins, A. Wheeler, P. Whitehead, N. Wilimovsky, A. B.
Williams, L. Wood, B. L. Yeager, P. Yuschak. H. Zadoretsky. B. J. Zahuranec.
Illustrators deserve special praise and thanks. B. B. Washington illustrated a large majority of the specimens. Other
illustrators include G. Mattson who served in this capacity with Dr. Ahlstrom for many years. B. Y. Sumida and H. Orr
at the Southwest Fisheries Center. B. Vinter at the Northwest and Alaska Fisheries Center and J. C. Javech at the
Southeast Fisheries Center. The original illustrations are archived at the Southeast Fisheries Center. Miami and Southwest
Fisheries Center, La Jolla.
During the final editorial processes, J. C. Javech and B. B. Washington mounted illustrations and remade many that
were of marginal quality. C. Wolf coordinated and reviewed the literature cited section and P. Fisher typed the literature
cited section as well as all last minute editorial changes.
The Editorial Committee:
H. G. Moser, Editor in Chief
W. J. Richards, Managing Editor
D. M. Cohen
M. P. Fahay
A. W. Kendall, Jr.
S. L. Richardson
CONTENTS
Welcoming Address. By /. Barrett vii
Frontispiece— Photograph of Elbert Halvor Ahlstrom. By /. R. Dunn viii
Dr. Ahlstrom. By R. Lasker _ ix
Introduction
Ontogeny, Systematics and Fisheries. By J. H. S. Blaxter __ __ 1
Ontogeny, Systematics and Phylogeny. By D. M. Cohen 7
Early Life History Stages of Fishes and Their Characters. By A. W. Kendall, Jr.. E. H. Ahlstrom and H. G.
Moser 1 1
Techniques and Approaches
Early Life History Descriptions. By E. M. Sandknop. B. Y. Sumida and H. G. Moser _ 23
Synopsis of Culture Methods for Marine Fish Larvae. By J. R. Hunter 24
Identification of Fish Eggs. By A. C. Matarese and E. M. Sandknop 27
Identification of Larvae. By H. Powles and D. F. Markle 31
Illustrating Fish Eggs and Larvae. By B. Y. Sumida, B. B. Washington and W. A. Laroche 33
Clearing and Staining Techniques. By T. Potthoff 35
Radiographic Techniques in Studies of Young Fishes. By /. W. Tucker, Jr. and J. L. Laroche 37
Histology. By y. y. Govoni _ 40
Scanning Electron Microscopy. By G. W. Bochlert _ _ 43
Developmental Osteology. By J. R. Dunn 48
Otolith Studies. By E. B. Brothers __ 50
Preservation and Curation. By R. J. Lavcnhcrg. G. E. McGowen and R. E. Woodsum 57
Development and Relationships
Elopiformes: Development. By W. J. Richards 60
Notacanthiformes and Anguilliformes: Development. By P. H. J. Castle _ 62
Elopiformes, Notacanthiformes and Anguilliformes: Relationships. By D. G. Smith 94
Ophichthidae: Development and Relationships. By M. M. Leiby 102
Clupeiformes: Development and Relationships. By M. F. McGowan and F. H. Berry 108
Ostariophysi: Development and Relationships. By L. .A. Fuiman 126
Gonorynchiformes: Development and Relationships. By H'. J. Richards 138
Salmoniforms: Introduction. By H '. L. Fink _ 1 39
Esocoidei: Development and Relationships. By F. D. Martin 140
Salmonidae: Development and Relationships. By A. W. Kendall. Jr. and R. J. Behnke 142
Southern Hemisphere Freshwater Salmoniforms: Development and Relationships. By R. M. McDowall _... 150
Osmeridae: Development and Relationships. By M. E. Hcarnc 153
Argentinoidei: Development and Relationships. By E. H. .Ahlstrom. H. G. Moser and D. M. Cohen 155
Stomiatoidea: Development. By A'. Kawaguchi and H. G. Moser 169
Stomiiforms: Relationships. By W. L. Fink 1 8 1
Families Gonostomatidae, Stemoptychidae. and Associated Stomiiform Groups: Development and Relation-
ships. By E. H. .Ahlstrom. \V. J. Richards and S. H. Wcitzman 184
Giganturidae: Development and Relationships. By R. K. Johnson _ 199
Basal Euteleosts: Relationships. By W. L. Fink _ 202
Myctophi formes: Development. By M. Okiyama _ 206
Myctophidae: Development. By H. G. Moser. E. H. Ahlstrom and J. R. Paxton 218
Myctophidae: Relationships. By J. R. Pa.xton. E. H. Ahlstrom and //. G. Moser 239
Scopelarchidae: Development and Relationships. By R. K. Johnson 245
Evermannellidae: Development and Relationships. By R. K. Johnson 250
Myctophiformes: Relationships. By M. Okiyama 254
Gadiformes: Overview. By D. M. Cohen 259
Gadiformes: Development and Relationships. By M. P. Fahay and D. F. Markle _ __ 265
Gadidae; Development and Relationships. By J. R. Dunn and A. C. Matarese 283
Bregmacerotidae: Development and Relationships. By E. D. Houde 300
Ophidiiformes: Development and Relationships. By D. J. Gordon. D. F. Markle and J. E. Olney 308
Lophiiformes: Development and Relationships. By T. W. Pietsch _ 320
Ceratioidei: Development and Relationships. By E. Bertelsen __ _ _ 325
Atherinomorpha: Introduction. By B. B. Collette 334
Beloniformes: Development and Relationships. By B B. Collette, G. E. McGowen. N. V. Parin and 5. Mito 335
Atheriniformes: Development and Relationships. By B. N. White, R. J. Lavenberg and G. E. McGowen 355
Cyprinodontiformes: Development. By K. W. Able 362
Lampriformes: Development and Relationships. By J. E. Olney 368
Mirapinnatoidei: Development and Relationships. By E. Bertelsen and TV. B. Marshall 380
Beryciformes: Development and Relationships. By M. J. Keene and A'. A. Tighe 383
Zeiformes: Development and Relationships. By A'. A. Tighe and M. J. Keene 393
Gasterosteiformes: Development and Relationships. By R. A. Fritzsche 398
Scorpaeniformes: Development. By B. B. Washington, H. G. Moser, W. A. Laroche and W. J. Richards 405
Cyclopteridae: Development. By A'. W. Able, D. F. Markle and M. P. Fahay 428
Scorpaeniformes: Relationships. By B. B. Washington, W. N. Eschmeyer and K. M. Howe 438
Tetraodontoidei: Development. By J. A/. Lets 447
Balistoidei: Development. By A. Aboussouan and J. M. Lets _ 450
Tetraodontiformes: Relationships. By J. M. Lets - 459
Percoidei: Development and Relationships. By G. D. Johnson _ 464
Serranidae: Development and Relationships. By A. W. Kendall, Jr. _ 499
Carangidae: Development. By W. A. Laroche, W. F. Smith- Vaniz and S. L. Richardson 510
Carangidae: Relationships. By W. F. Smith- Vaniz 522
Mugiloidei: Development and Relationships. By D. P. de Sylva 530
Sphyraenoidei: Development and Relationships. By D. P. de Sylva 534
Polynemoidei: Development and Relationships. By D. P. de Sylva 540
Labroidei: Development and Relationships. By W. J. Richards and J. M. Leis 542
Acanthuroidei: Development and Relationships. By J. M. Leis and W. J. Richards 547
Blennioidei: Introduction. By R. H. Rosenblatt 551
Schindlerioidei: Development and Relationships. By W. Watson. E. G. Stevens and A. C. Matarese 552
Trachinoidea: Development and Relationships. By W. Watson. A. C. Matarese and E. G. Stevens _ 554
Notothenioidea: Development and Relationships. By E. G. Stevens. W. Watson and A. C. Matarese — 561
Blennioidea: Development and Relationships. By A. C. Matarese. W. Watson and E. G. Stevens 565
Ammodytoidei: Development and Relationships. By E. G. Stevens. A. C. Matarese and W. Watson 574
Icosteoidei: Development and Relationships. By A. C. Matarese. E. G. Stevens and W. Watson 576
Zoarcidae: Development and Relationships. By M. E. Anderson _ 578
Gobioidei: Development. By D. Ruple 582
Gobioidei: Relationships. By D. F. Hoese 588
Scombroidei: Development and Relationships. By B. B. Collette, T. Potthoff, W. J. Richards, S. Ueyanagi,
J. L. Russo and Y. Nishikawa 59 1
Stromateoidei: Development and Relationships. By M. H. Horn _ 620
Gobiesociformes: Development and Relationships. By L. G. Allen 629
Callionymidae: Development and Relationships. By E. D. Houde 637
Pleuronectiformes: Development. By E. H. Ahlstrom, K. Amaoka, D. A. Hensley, H. G. Moser and B. Y. Su-
mida 640
Pleuronectiformes: Relationships. By D. A. Hensley and E. H. Ahlstrom - 670
Literature Cited _ 688
Index _ - 746
Photograph of Symposium Attendees 760
VI
Welcoming Address
IzADORE Barrett
Director of the Southwest Fisheries Center
ON behalf of the National Marine Fisheries Service's Center Directors, sponsors of the Symposium on the Ontogeny
and Systematics of Fishes, I am pleased and honored to welcome you to La JoUa. We are here to honor the memory
of an outstanding biologist, Elbert Halvor Ahlstrom, known to his friends and colleagues as Ahlie, and his contributions
to fisheries science.
As fishery biologists we all recognize the vital importance and contributions of systematics and students of evolution
to the development of fishery science. Less well known or appreciated is the unique role and interrelationship of the
early life history studies of fishes and the assessment of the role of ontogenetic characters in fish systematics. This was,
of course, the field of fisheries research to which Ahlie dedicated 40 years of his professional life and where he initially
evolved the special methods and techniques which have so greatly influenced the work of fishery biologists around the
world.
I know that I speak for the Directors of the four fisheries centers— the Northwest and Alaska Fisheries Center in
Seattle, the Southwest Fisheries Center in La Jolla, the Northeast Fisheries Center in Woods Hole, and the Southeast
Fisheries Center in Miami when I say that I am proud that the National Marine Fisheries Service is the sponsor of this
symposium. I believe that this gathering will be a landmark in fisheries science, a unique event which has brought
together eminent scientists from 10 countries to present 87 papers reviewing the major fish groups, with particular
attention to ontogenetic characters and their utility in assessing phylogenetic relationships. I fully anticipate that the
resulting symposium volume which will be based on the papers presented here will stand as a definitive work in larval
fish biology for many years to come.
Again, a warm welcome to all of you and especially to Marge Ahlstrom who is seated in the audience this morning.
I hope that the weather and circumstances will cooperate and that your stay here in one of the most attractive cities of
the United States will be pleasant and productive.
P.O. Box 271, La Jolla, California 92038.
Dr. Ahlstrom
Reuben Lasker
MY colleagues have entrusted to me the pleasant task and distinct privilege of saying a few words in remembrance
of Dr. Elbert H. Ahlstrom, to whom this symposium is dedicated. Like most of you I was his colleague for many
years, 23 to be exact. He was also my friend and mentor to whom I could go when I needed advice and where I knew
I would be heard as an individual with the bond of common scientific endeavors.
For those of you who did not know Dr. Ahlstrom 1 would like to capsulize his enormous contribution to systematics
and fishery science by outlining what I believe to be his major scientific contributions. Ahlie realized in the late 40's
that the study of eggs and larvae could give us information about fish populations unobtainable from fishery statistics,
the mainstay of fishery science at that time. He believed, rightly, that the ease with which eggs and larvae could be caught
allowed an assessment of the geographic distribution and the seasonal extent of spawning of pelagic species. He recognized
that any assessment of a fish population was dependent on surrounding that population in time and space and that this
would require a major effort. He was the first. I believe, to determine the extent of a major pelagic fish population using
this technique.
The simplicity and thoroughness of the plankton net made an impression on him and, while he sought to improve
collecting techniques constantly, he consistently analyzed the errors of the plankton net so that this tool could be used
more and more reliably. Today, it is still one of the most powerful collecting and assessment tools we have, largely
because of his diligence and persistence.
The scope and thoroughness of Dr. Ahlstrom's work was particularly important. His taxonomic skills are attested to
in the many papers he wrote and which stand today as mainstays of the systematic and fishery literature. He liked to use
the title "Kinds and abundance of fishes" and usually provided taxonomic lists in these of several pages in length. His
point, of course, was to detail the complexity and uniqueness of particular oceanic regimes and to set the ground work
for ecological research which inevitably followed.
Well, what of his other attributes? I used to call him the modem Renaissance Man because I realized whenever I had
occasion to meet him socially that he knew almost all there was to know about the arts and the sciences. Of his fabulous
classical record collection 1 recall that 1 asked him once if he really listened to all of them. His reply was "we used to
hear each one once a year, but now, since the collection has grown so large, it's once every two years." He belonged to
the San Diego Great Books Society, and read them all. Engage him in conversation and you would find out quickly he
knew literature, fine wines, photography and baseball, to name a few. I would like to sum up this brief eulogy by pointing
out an example of one aspect of Ahlie which holds my greatest admiration: that is, his dedication to work. One incident
during our relationship illustrates the point I wish to make.
When Science Fairs started to become the vogue in San Diego, Dr. Ahlstrom was asked to host a group of young
Science Fair participants to teach them something about oceanography. He arranged to take out the old Bureau of
Commercial Fisheries ship, the Black Douglas, for a day to illustrate collecting methods at sea. In fact, the day was
beautiful, but there was a swell upon the sea and no sooner did we get out of the harbor than almost everyone, except
Ahlie and some of the seasoned veterans, felt the effects of a rather pronounced roll for which the Black Douglas was
famous, even in the calmest of seas. Dr. Ahlstrom proceeded with his typical dedication to illustrate Nansen bottles,
plankton nets, and bathythermographs to the group of Science Fair students who were becoming less and less interested
and more and more seasick.
Ahlie continued with a single-mindedness of purpose and a dedication that was so characteristic of him. Without his
noticing, a caucus was held by these young students and a representative meekly asked, "Dr. Ahlstrom, may we please
go home?"
Two versions of what happened next were told to me later. The first was that Ahlie responded immediately to the
problem and ordered the ship to port. Another version was that Ahlie continued until he was finished, made sure he
had a proper sample, and then ordered the ship into port. I'm afraid I can't tell you which is correct— I was in a bunk,
seasick! I meant this story as a small illustration of Dr. Ahlstrom's dedication to his work.
He was a dedicated scientist who had an insatiable curiosity about the biotic world and who was convinced that what
he was doing was important and would advance fishery science. This symposium is one piece of evidence that he was
right.
Now the question must be asked— how is it that Ahlie could be so dedicated to work and yet have found time to
become a true example of a Renaissance man, with a deep knowledge of art, wine, architecture, photography, sports,
and much more? I pondered this with admiration for many years and I think I have the answer. He was one of those
rare individuals who never cease learning, because he had a true scholar's love for learning. I like Robert Whittenton's
description of Sir Thomas More when I think of Ahlie: he was, like More, "a man for all seasons."
Southwest Fisheries Center, P.O. Box 271, La Jolla, California 92038.
Photograph of Elbert Halvor Ahlstrom, by J. R. Dunn.
INTRODUCTION
Ontogeny, Systematics and Fisheries
J. H. S. Blaxter
IN the inter-war years work on fish eggs and larvae was Umited
to studies on horizontal and vertical distribution with a view
to completing our knowledge of the early life history of different
species. Resources for research were then much more limited
than they are today and most work was done on the important
food fishes. In the 1 950's a great expansion took place as fisheries
biologists realised how much a study of early life history would
be a key to solving some of their problems. This expansion took
place on a broad geographical and mtemational front, but great
credit must be given to the foresight and imagination of E. H.
Ahlstrom. who built up a team of biologists at La Jolla who
then and subsequently, played a major role m leading and de-
veloping this field with special reference to the fisheries of the
California Current.
In the last two decades the output of publications has risen
at an exponential rate as evidenced, for example, by the 62
papers in the 1973 Early Life History Symposium held in Oban
(Blaxter, 1974) and the 139 papers in the 1979 Symposium at
Woods Hole (Lasker and Sherman. 1981). Furthermore, in a
selected hibhography of pelagic fish and larva surveys prepared
by Smith and Richardson (1979), some 1200 papers are listed,
most of them published in the last 30 years. Ahlstrom was
certainly a major catalyst in this reaction, but it is sad to record
that his obituary appeared in the Proceedings of the 1979 Sym-
posium, although he was still alive and present at the meeting
itself to impart his wisdom and expertise.
It is proposed to discuss the post-war advances in our knowl-
edge of early life history stages under five headings: (1) as they
impinge on systematics and taxonomy. (2) the success and role
o{ experimental work in tanks and of modelling, (3) the scaling-
up of tank studies to large enclosures and embayments, (4) the
application oi sea surveys to test models, to investigate the stock-
recruitment relationship and to measure spawning stock bio-
mass, and (5) \he future.
Systematics and Taxonomy
A number of techniques have been developed to help in the
identification and classification of fish larvae. Since the devel-
opment of the skeleton and meristic characters are now so im-
portant in identification, techniques of clearing and staining or
x-radiography have become standard methods for examining
the internal osteology of larvae (Ahlstrom and Moser, 1981).
Morphometries and body pigmentation are also important and
are used extensively by Russell (1976) in his monograph on fish
eggs and larvae of the N.E. Atlantic.
Rearing experiments have shown that the sequence of de-
velopmental events may also be specific in character. For ex-
ample the development of the acoustico-latcralis system and
swimbladder in herring as shown by Allen, Blaxter and Denton
(1976) is a long-drawn-out affair and quite different from that
of the larval anchovy as described by O'Connell ( 1 98 1 a) or the
menhaden or sprat. There are several larval features, such as
the swimbladder and other internal organs, or features of the
labyrinth, which would help in the separation of similar-looking
species if only they were not obscured by fixation.
Often the taxonomist (or fisheries biologist) resorts to count-
ing menstic characters such as vertebrae, fin rays, scales or gill
rakers. Yet many of these characters have been shown by ex-
periment to be labile and to respond to environmental condi-
tions during early development. The earlier work, mainly on
freshwater species such as the sea trout, was summarised by
Taning (1952). Since then a range of further studies by Fahy,
Lindsey (e.g., see Fahy, 1982) and others have confirmed the
earlier experiments, showing that temperature, salinity and oxy-
gen level influence meristic counts and that there is a critical
period when this influence operates. Little work has been done
on marine species although Hempel and Blaxter (1961) showed
that temperature and salinity both influence myotome and ver-
tebral counts in herring (the species in which stock separation
by meristic counts has been most widely applied).
It seems likely that any environmental variable which influ-
ences the relationship between differentiation and growth will
affect the meristic count by determining the amount of embry-
onic tissue which is present when the differentiation into skeletal
units lakes place. The larval taxonomist needs to be cautious
in interpreting small differences in meristic values, especially
when they are related to clines or other types of geographical
distribution. That is not to say, however, that there is no un-
derlying genetic mechanism. The environment acts as a "fine-
tuning" mechanism. Whether this fine-tuning is accidental or
adaptive might well be worth discussion at the symposium.
A warning also needs to be directed at morphometries. Rear-
ing experiments in different-sized tanks by Theilacker ( 1 980b)
show the influence of space on growth rates. Compansons of
reared and wild fish larvae, especially of herring by Blaxter
(1976), show that tank-reared fish are often shorter and fatter
than their wild counterparts at the same developmental stage.
There seems to be an interplay between diet and activity which
is enhanced by the confinements of the rearing tank. This makes
it difficult to extrapolate growth criteria from tanks, such as
condition factor, to establish, for example, the nutritional status
of larvae at sea (Fig. 1).
A further and serious problem identified by the handling and
use of live larvae is the shrinkage caused by capture and fixation.
A number of workers such as Blaxter (1971), Schnack and Ro-
senthal (1978), Theilacker (1980a) and Bailey (1982) have ad-
dressed this problem but the most significant findings are those
of Hay (1981) on Pacific herring. Feeding larvae from rearing
experiments were released into the mouth of a plankton net at
sea and then fixed by various techniques after capture. Shrinkage
in body length ranged from a mere 5% to a massive 43% de-
pending on the technique. Extensive voiding of gut contents also
occurred. The implications of these results in morphometric or
feeding studies will not be lost on the present audience.
1
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
•20
.16-
(J
•12
•08-
HERRING
WILD
12 16
LENGTH (mm)
"20
Fig. 1 . Comparison between range of condition factors (C.F.) as dry
weight/length^ of wild herring caught at sea by plankton net and reared
herring larvae near starvation (from Blaxter, 1976).
Finally, the ageing of larvae by daily ring formation in the
otoliths should be mentioned. This technique was pioneered by
Brothers et al. (1976) on anchovy larvae and California grunion
following Pannella's suggestion that daily increments were being
laid down in the sagittae of some temperate fish species. The
findings were validated by rearing larvae in tanks and sampling
the population at intervals of 1-7 days. Struhsaker and Uchi-
yama (1976) supported these results from their work on the
Hawaiian nehu and subsequently the technique was widely
adopted in fisheries laboratories. Attempts by Geffen (1982) to
manipulate ring formation in cod, herring, plaice, salmon and
turbot larvae by varying the photoperiod, temperature and feed-
ing regimes did not lead to any consistent result — the ring de-
position was frequently not daily and the main determinant in
herring and turbot seemed to be growth rate— the higher the
growth rate, the higher the rate of ring deposition. Bailey (1982),
however, found otolith rings deposited daily over a 10-day pe-
riod in post yolk-sac Pacific hake larvae reared in tanks. Sea-
caught larvae with more than about 30 increments were less
satisfactory because of the appearance of different types of ring
and it was not certain whether they were daily. Dale (1984) in
a recent study of reared Atlantic cod otoliths using electromi-
croscopy, found daily rings in a 12L/12D cycle but not in the
dark. Daily ring deposition only continued, however, for a few
days post-hatching.
Although the ageing of anchovy and grunion from daily rings
seems reliable, further validation experiments are required at
sea. This is conceptually difficult on a wild stock of larvae of
mixed age and it is notoriously difficult to remain over a single
population of larvae for many days. Mass release of reared
larvae into the sea remains an ambitious possibility. Perhaps
best of all such a release should be into some large enclosure
system initially free of a larval population. Validation experi-
ments must also test the more unusual environmental condi-
tions which apply in high latitudes where, for example, daylight
prevails over the full 24 hours.
Experimental Work
The functional anatomy approach to taxonomy so elegantly
described in a recent review by Moser (1981) shows the extent
to which structure can be used to deduce function. The inter-
action of this approach with that of the experimentalist has
yielded much useful information.
Since the 1950's increasing success in rearing marine fish
larvae may have provided the taxonomists with help as well as
some doubts as described in the last section. It has also led to
a wide literature on the physiology, behaviour and physiological
ecology of larvae (and the use of larvae in pollutant bioassay)
as biologists seized the opportunity to exploit such new and
valuable material. Perhaps the most credit should be given to
Shelboume (1964) for his extensive and painstaking rearing ex-
periments on plaice, and later sole, at Port Erin, Isle of Man.
These experiments undoubtedly led to the present wide practice
of marine finfish aquaculture with the expanding commercial
use of turbot, sole, bass, bream and gilthead.
Rearing may still be considered as something of an art and
is often most successful in the hands of dedicated people with
a "feel" for what is right or wrong. Undoubtedly a breakthrough
was made in finding suitable food for larvae. It is significant
that both plaice and sole can take Anemia nauplii from first
feeding as can some races of herring. This resulted in another
U.K. focus for rearing at Aberdeen, and later Oban, developed
by Blaxter (1968) on the herring. Species with smaller larvae
(with smaller mouths) were only successfully reared when Las-
ker's group at La Jolla (Lasker et al., 1970; Theilacker and
McMaster, 1971; Hunter, 1976) developed the use of the rotifer
Brachionus plicatilis and the naked dinoflagellate Gymnodmium
splendens as small food items for early-stage larvae of species
like northern anchovy and jack mackerel. About the same time
Howell (1973) also used Brachionus to rear turbot larvae at Port
Erin.
Subsequently a number of factors have been identified to add
to our corpus of knowledge on rearing. These include the need
for good water quality, with the interesting idea of "green water"
culture of larvae in fairly high densitiesofC/j/oreZ/a which seems
to damp out fluctuations in metabolites, and perhaps enhance
oxygenation as well as providing secondary feeding for the larvae
(e.g., Houde, 1977; Morita, 1984). Adequate light for visually-
feeding larvae and the need to prevent excessive bunching of
larvae or their prey are also important, as is the quality of the
food. Success or failure may now depend on the fatty-acid profile
of the Anemia nauplii which are still used by most workers in
the later stages of rearing. Artificial diets of encapsulated or
particulate food are also being developed but have yet to be
introduced as a standard technique for early rearing.
Before turning to the extrapolation and application of exper-
imental data to modelling, mention must be made of Haydock's
(1971) and Leong's (1971) work on the induction of spawning
in the croaker and anchovy by pre-treatment with an appro-
priate photoperiod followed by hormone injection. This has
been applied subsequently to the menhaden by Hettler (1981),
and to many other species, and has become a standard method
for workers requiring eggs over long periods or at a specific time.
We now have the widest knowledge of the development, be-
haviour and physiology of both anchovy and herring larvae (see
Fig. 2) but there are several species such as cod, jack mackerel,
mackerel, plaice and turbot which run them a close second.
BLAXTER: ONTOGENY, SYSTEMATICS, FISHERIES
Lateral line
Respiration
Red muscle
Reynolds number
(Re) and
hydrodynamic
Viscous
regimes
Digestive tract
Re<lO
Time to 50%
starvation
Larval period
First feeding
"^ 1
Photopic vtslon I
Lens retractor muscle
Improved accommodation^ — "
^ *-'-
Threshold
tof
schooling
Initial swim bladder Inflation
Olel vertical movements
, ---^T-
Functional eye First rods
Increase in number of neuromasts
Scotopic vision
Rod recruitment continues
Many rods
Canal formation
First Epidermis
RBC's thickens
Many
RBC s
Cutaneous respiration
Superficial
1 layer
Transition,
"?0<R'e'<2°°
Functional gut
Movable lower Jaw
2.5 days
:t
I I
5
Gill respiration
Scale formation
Midline
2-3 layers 3-4 layers
7-8 layers
Re>200
Stomach forms
.It-
Expandable mouth
3.3 days 4 days
T^ 1 1 1 — I 1 1 1 1 1 1 1 1 1 1 1 — 1-
15 20 25 30
Length (mm)
Juvenile
period
Filler feeding
15 days
*
1 — I 1 1 1 1 1 —
2.5 5 10 15 20 25 30
35 40 45
Days at 16° C
— 1 1 1 1 1 1 1 —
50 55 60 65 70 75 80
Fig. 2. Events during development of the northern anchovy. RBC = red blood cells. Time to 50% starvation is number of days to starvation
at which 50% of the fish died (from Hunter and Coyne. 1982).
Much of this work is summarised by Theilacker and Dorsey
(1980).
Over the past few years the assembly of much basic data has
allowed the current vogue for modelling to be applied to fish
larvae. Modelling is an attempt to synthesise and simplify basic
data usually in mathematical form. Mathematical models are
often iterative and they have the value of being in a form suitable
for computers. Laurence (1981) has recently reviewed modelling
work on fish larvae and the complexity and type of interaction
is shown in Fig. 3. The main problem addressed has been that
of feeding. The earlier models of Blaxter (1966), Rosenthal and
Hempel ( 1 970), Blaxter and Staines ( 1 97 1 ) and Hunter (1972)
estimated the feeding efficiency of larvae, the volume of water
searched in unit time and the density of food required to give
good survival and growth. More sophisticated models have now
been developed (e.g., Jones and Hall, 1 974; Beyer and Laurence,
1981) and Vlymen's (1977) model allows for the prey species
being non-randomly distributed.
The need for larvae and their prey to co-exist temporally was
spelled out by Gushing ( 1 975) in his match-mismatch hypothesis.
Thus the timing of reproduction appears to have evolved to
synchronise the larval stages with the main phase of the annual
production cycle. Spawning is probably controlled in most tem-
perate fish species by photoperiod and temperature which are
not the only determinants of plankton production. Hence a
match or mismatch is possible between this production and the
presence of fish larvae with a resulting influence on year class
strength.
An early paradox existed in that the density of the larger
micro-zooplankton such as copepod nauplii required for good
growth and survival in tanks was of the order of 1 organism/
ml. Such densities are rarely found in the sea as judged from
normal plankton sampling. This led to the suggestion of micro-
scale patchiness of food in the sea, which might occur at inter-
faces such as steep thermoclines and at tide- and wind-induced
fronts. The integrity of such microscale patchiness would not,
of course, be obvious using nets sampling large volumes.
This led Lasker (1975) to bioassay samples of water taken at
different depths and places off the Califomian coast, using an-
chovy larvae both hatched and tested on board ship. Chloro-
phyll-rich layers with very high densities oi Gymnodinium were
found near the thermocline. The bioassay showed good larval
feeding in these water samples, suggesting that patchiness, in-
deed, might be a valid concept. This was to some extent con-
firmed by later findings that stable weather conditions (which
maintained the thermocline) favoured good year classes of an-
chovy larvae off the Califomian coast (Lasker, 1981). Owen
( 1 980) has subsequently shown from samples taken by plankton
pumps and water bottles that patchiness of microzooplankton
such as copepod nauplii and tintinnids and various protozoan
species and phytoplankton (some of which are known to be the
food of anchovy larvae) exist off the Peruvian and Califomian
coasts on the scale of a few centimetres up to one metre (see
Fig. 4). Only Houde and Schekter ( 1978) have attempted to rear
larvae in simulated food patches and found that survival of sea
bream was similar when they were exposed to 3 h of food per
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
PHYSICAL A CHEMICAL INFLUENCES
CIRCULATION
DIFFdSlON
80UNDARV EXCHANGE
DISCONTINUITY EVENTS (STORMSj
PRIMARY PRODUCTION &
ORGANIC RECYCLING
POLLL TION OR TOXICITY
ABIOTIC FACTORS (TEMPERATURE
SALINITY. OXYGEN)
IMPORTANT I ARVAL FISH INTERACTIONS
LARVAl
PREDATORS
MOSTLY
UNIDENTIFIED
RECRUITMENT
ASSESSMENTS
USHERY MODELS
ECOSYSTEM MODELS
MANAGEMENT STRATEGY
LARVAL MORTALITY DETERMINED DIRECTLY IN LINKAGES A A B
o
o
tl'RREMLY rWPHASlZED STL:nrFS
STtlDIES FOR EMPHASIS
SI I DthS ME IMMFniAlE
I ESSEk IMPORT ASCE
DIRECl HACnONAI IINKAGES
I ISKAGES OE I ESSER IMPORTANCE
Fig. 3. A generalised scheme for the main interactions between larval fish and their biotic and abiotic environment, providing a basis for
modelling (from Laurence, 1981).
day as when fed at the same food level continuously. Clearly,
expeiiments need to be devised to test the effect of spatial rather
than temporal food patchiness.
The evidence is thus accumulating, but very slowly, that lai^al
survival may depend on the extent and stability of microscale
food patches or interfaces, at least in some areas. It may be that
the rather high food densities required in small-scale tank rear-
ing do indeed apply to conditions in the sea and that such
densities are only found in patches.
SCALING-UP
Two major areas may be identified where rearing work has
been extended into large-scale containers. The first of these are
the large onshore enclosures and embayments used by the pres-
ent generation of Norwegian biologists; the second are the deep-
water plastic bags used by Scottish workers in Loch Ewe on the
Scottish West Coast. The Norwegians have achieved remarkable
growth and survival rates for herring and cod larvae, as high as
30-70% survival from hatching to metamorphosis, in shallow
4,000-60,000 m^ enclosures (Oieslad and Moksness, 1981;
Kvenseth and Oiestad, 1984). The Loch Ewe bags, which are
deep cylinders, of about 300 m\ have been used for rearing
herring and cod, but with much less success than the Norwegians
(Gamble et al., 1981; Gamble and Houde, 1984).
Possibly volume itself is important, or more likely the ratio
between volume and wall area. The interface between wall and
sea water is not a natural one for fish larvae, feeding may be
difficult al the interface, and food may aggregate there in an
inaccessible form. Morita (1984) reports that Pacific herring
larvae have recently been reared in 20 m' tanks with a 46%
survival from hatching to a mean length of about 7 cm in 1 1 2
days. This spectacular result may have been partly a feature of
a fairly large onshore tank but also the "green water" technique
mentioned earlier. Hunter (1984) suggests that the high survival
in some large tank or enclosure experiments is achieved by the
elimination of predators. To the present author a combination
of optimal feeding conditions and low predation seems to be
the likely cause.
The events have been described so far in a topsy-turvy way,
in that sea surveys have always been the most widely-adopted
approach to problems associated with the early life history of
fish. The experimental and enclosure studies are the icing on
the research cake, although both Norwegian and Japanese work-
ers are seriously considering the possibility of restocking de-
pleted inshore fisheries or topping-up poor year-classes of cod
and herring by releasing reared late-stage larvae or O-group
juveniles.
Sea Surveys
These are expensive in terms of ship-time and manpower.
Originally designed to advance our knowledge of spawning
grounds, larval drift, and horizontal and vertical distribution,
they are often now linked to more practical aims. Nevertheless,
superb time-series exist for areas like the California Current and
North Sea as a result of the patience and foresight of earlier
workers like Ahlstrom and later workers like Smith and Saville
(see review by Smith and Richardson, 1977). Sea surveys have
always been a rich ground for innovative science, in terms of
sampling techniques, interpretation and usage. Experimenters
and modellers have provided a great boost for this work, allow-
ing new interpretations to be made and new hypotheses to be
tested.
No more mention will be made of the matrix-filling role of
sea surveys— namely the completion of details of life history,
which is still taking place and has been much aided by the vast
improvement in egg and larval identification in the past two
BLAXTER: ONTOGENY, SYSTEMATICS, FISHERIES
CONCENTRATION (no/X)
20 0
D 1000 2000
3000
<■
^
X
J
20 4
"-^^
J
e
Prorocentrum — ^ ^ — ■C' '
_-^-^
^
I
?0 8
J^^ \ —
-Nit/schia
1—
— "" " " ^ — \
a.
- < J^
UJ
Q
212
f^"^'^"^"'---
V L____
"jt
216
PHYTOPLANKTON
25
20 0
20 4
20 8
21 2
CONCENTRATION (no /i)
50 75
100
216-
Fig. 4. Vanation in concentration of microplankton in samples from 20 cm depth intervals in the chlorophyll maximum layer over the coastal
shelf of the Southern California Bight dunng March. 1976. Prorocentrum, tintinnids and copepod nauplii are all food items for larval anchovy
(from Owen. 1980).
-
• •r-
)
Tintinn
r
t
{ Nauplior
s-^r copepods
/
- — Noctiluca
MICRO-ZOOPLANKTON
decades. Improvement in plankton nets and young fish trawls
means that vertical profiling and quantitative sampling have
finally come-of-age. This ability to sample quantitatively is the
single most important advance in allowing larval populations
to be assessed reliably and for allowing models to be tested. The
outcome is two-fold. The door is open for biomass estimates of
spawning stock from egg and larval surveys and for testing the
possible factors in the stock-recruitment relationship. Each of
these will be considered in the final part of this paper.
/. Biomass estimation. — ¥or many years population dynami-
cists lacked good information on the absolute size of the spawn-
ing stock and regulation was largely achieved by minimum mesh
and landing sizes. Of late, as a result of catastrophic declines in
some species, whole fisheries have been closed or controlled by
quotas and total allowable catch (TAC). The use of TAC's has
been greatly aided by virtual population analysis and also by
sonar-based fish counting surveys; these give an estimate of total
stock size, the reliability of which depends on the extent of the
survey, the ability to identify the species in question and the
precision of the calibration of target strength.
To supplement the results, estimates of spawning stock size
have been made on an ad hoc basis by counting eggs and larvae
and converting them into the parental spawning stock biomass
by a knowledge of fecundity, age distribution and sex ratio. Some
of the pioneering work was done by Sette and Ahlstrom (1948)
on Califomian pilchard and Simpson (1959) on North Sea plaice.
Saville, Baxter and McKay (1974) counted the demersal eggs of
the herring on the small spawning ground of Ballantrae Bank
in the Clyde. This was later extended by Saville and McKay
(see Saville, 1981) to herring larval surveys in the North Sea
and off the Scottish west coast. The biomass of Pacific hemng
is now routinely assessed from the intertidal egg deposition along
the coast of Canada and the USA as described in the recent
Nanaimo Herring Symposium (Hay, 1984; Haegele and
Schweigert, 1984). Similar, but ad hoc. data are available for
the northern anchovy from the work of Smith (1972), Parker
( 1 980) and Picquelle and Hewitt (1983), for the Atlantic mack-
erel from Lockwood, Nichols and Dawson (1981) and Berrien,
Naplin and Pennington (1981) and for North Sea cod from Daan
(1981). Some of these data give absolute measures, some relative
ones from year-to-year, often related to biomass estimates by
other means.
This survey technique has notable disadvantages. It must be
done at a limited time of year and is obviously easiest to interpret
for one-off spawners. The survey must be done rapidly and as
near the spawning season as possible to overcome any errors
caused by mortality between spawning and sampling. Although
it can be applied to a closed fishery, the age structure of the
population is required to compute the aggregate fecundity, hence
scientific sampling of the adults is required.
2. Stock-recruitment.— The relationship between the size of the
spawning stock in any year and the number of recruits it supplies
to the fishery subsequently is vital information for the regulation
of fisheries. This is specially true where recruitment overfishing
is prevalent as in the clupeoids. Over many years a stock-re-
cruitment relationship may be obtained empirically in any fish-
ery, but this is time-consuming and usually contains inexplicable
features. While, as might be expected, low spawning stock leads
to low recruitment, high spawning stocks may also give unex-
pectedly low recruitment, as the result of density-dependent
effects. Alternatively spawning stocks of a given size can yield
enormously different brood strengths, of the order of 10-100
times, in a quite unpredictable way.
It is not surprising that the underlying causes of the control
of brood strength are of much interest to fisheries biologists and
have received the attention of experimentalists and modellers.
Most marine fish have a very high fecundity, of the orders of
tens of thousands to a few million. From such a starting point
mortality must be very high and it is surprising that brood
strength variations are not even more variable than is actually
the case. What then do we know of the mortality rate of eggs
and larvae in the sea? Are there critical periods when it is es-
pecially high? What are the causes of mortality?
Hjort's original hypothesis, now some 70 years old, expressed
the view that a critical period existed after yolk resorption as
the larvae sought external food sources. This hypothesis was
supported by earlier rearing experiments in which very high
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
mortalities occurred at first feeding. Measurements of mortality
rates of eggs and larvae at sea tend to show a high but continuing
mortality of perhaps 5-20% per day. The results of sea surveys
are, however, often difficult to interpret because of the need to
sample within a discrete larval population over a long time.
May (1974), in his review of this subject, concluded that star-
vation at the end of the yolk-sac stage may often have a major
influence on brood strength but that mortality from fertilization
to the O-group stage is the ultimate determinant.
The results of modelling and the tests of the patchiness hy-
pothesis which have already been discussed support the idea
that first feeding is a critical time, although not having, neces-
sarily, the dominant effect claimed by Hjort. Experimenters and
modellers have also derived further concepts for testing. The
major sources of mortality are identified as starvation and pre-
dation. Starvation, of course, only operates from the end of the
yolk-sac stage. Blaxter and Hempel (1963) used the expression
"point-of-no-retum" to express the point at which larvae, as a
result of starvation, are too weak to feed even if food becomes
available. Sometimes called "ecological death" or "irreversible
starvation" this is a useful concept for assessing the chances of
larval survival under different conditions. For larvae in a good
nutritional state the time to the point-of-no-retum may be only
1-2 days in a small newly feeding larva like the anchovy, but
2-3 weeks in a well grown flatfish larva like the plaice (see
Theilacker and Dorsey, 1980). Implicit, also, in the concept is
that larvae can live for some time after the point-of-no-retum.
During this time they may be especially liable to capture by nets
and, without adequate knowledge, a false impression might be
obtained of the size or nutritional state of the larval population.
The assessment of nutritional state of larvae has been of wide
interest in recent years, in the hope of relating this to brood
strength. Initially Blaxter ( 1 965) measured the condition factors
of tank-reared herring larvae after varying periods of starvation
and then later compared the results with the condition factors
of sea-caught herring larvae (Blaxter, 1971). It was found that
most sea-caught larvae had much lower condition factors than
starving tank-reared larvae and it became apparent that the
extrapolation of tank criteria to the sea was invalid because the
tank larvae were short and fat compared with wild larvae (see
Fig. 1). This means that condition factor comparisons of wild
larvae are only valid on a relative basis from year-to-year or
place-to-place (e.g., Chenoweth, 1970; Vilela and Zijlstra, 1971)
and only then if one can be satisfied that shrinkage after capture
is consistent. The problems of tank ; sea comparisons and
shrinkage are unfortunately likely to be the most serious in long
clupeoid larvae to which these experiments have been applied.
No one has checked their validity in the more common type of
larvae with a shorter body form.
These problems led to work at Oban and La Jolla on histo-
logical criteria for assessing starvation (Ehrlich et al., 1976;
O'Connell, 1976; Theilacker, 1978). O'Connelfs work on an-
chovy larvae deserves special mention. He found from screening
the state of the body organs such as pancreas and gut that these
showed increasing signs of degeneration as starvation pro-
ceeded. On applying his criteria to sea-caught anchovy larvae
O'Connell (1981b) found evidence for quite a high percentage
of larvae suffering from advanced starvation and considerable
differences in the incidence of starvation in closely adjacent
areas. This method is now being applied by Theilacker on jack
mackerel larvae from year-to-year and is likely to be adopted
on a routine basis.
The other cause of mortality, predation, has recently become
fashionable following the work of Eraser, Lasker, Lillelund and
Theilacker and subsequently Kuhlmann, von Westemhagen and
Rosenthal, Bailey, Purcell and several other workers (See re-
views of Hunter, 1981, 1984). Copepods, euphausiids, amphi-
pods and chaetognaths are all implicated but perhaps medusae
are the most voracious group of predators (Bailey and Batty,
1983), especially for inshore spawners like Pacific herring. Pre-
dation, of course, operates from the moment of spawning and
Hunter and Kimbrell( 1980) and MacCall (1980), in particular,
have discussed the incidence of density-dependent cannibalism
of spawning anchovies on their own eggs and larvae. It is gen-
erally thought that strong selection pressure exists for fast growth
which will take larvae speedily through the more vulnerable
early stages. Larvae have been shown experimentally to be less
vulnerable when they are larger, their escape speeds are higher
and their recovery from a predator attack (for predators of a
given size) more likely. As Hickey (1979, 1982) has shown, an
efficient wound-healing mechanism exists, allowing larvae to
recover from bites, stings and other forms of damage. The high
survival rates of larvae reared in the absence of predators (Kven-
seth and Oiestad, 1984; Morita, 1984) suggest strongly that
predation is a major source of mortality in the sea. Although it
is difficult to assess the relative importance of starvation and
mortality in any larval population, it is also clear that the two
must interact in the sense that starving larvae will be more
susceptible to predation.
The Future
In this paper modelling has been only briefly discussed. The
method is now widely used for setting up hypotheses about
feeding, starvation, predation, cannibalism and other factors
associated with the stock-recruitment relationship and biomass
estimation. This approach is likely to continue as a basis for
sea surveys. It seems uncertain whether biomass will be routinely
estimated by egg and larval surveys except perhaps in Pacific
herring and northern anchovy. The cost is too high and sonar
surveys, if the problems can be ironed out, seem to be a better
bet.
Experimental data on predation still need to be collected and
few correlations exist between predator populations and egg and
larval mortality in the sea. In fact mortality studies on eggs and
larvae in the sea in general need to be perfected since the prob-
lems of following discrete populations and of ageing larvae are
still not fully solved. At least one source of information is largely
untapped and that is the explanation for the high survival rates
of larvae in large enclosures. In particular the distribution of
the larvae and their food in these enclosures is not known and
may throw light on the validity of the patchiness hypothesis.
Information on frontal systems, and interfaces as a result of tide,
wind, upwelling and thermo— and halo— clines is now quickly
being assembled by hydrographers and marine biologists. The
larval biologists should be ready to exploit the results.
It will be apparent to the audience how far research into the
early life history of fish has advanced in the last 30 years. A
major force has been the work off"the Califomian coast generated
by Ahlstrom and his recruits at La Jolla. It is therefore very
fitting that this symposium should be dedicated to his memory.
Scottish Marine Biological Association, Dunstaffnage
Marine Research Laboratory, P.O. Box 3, Oban,
Argyll, Scotland.
Ontogeny, Systematics, and Phylogeny
D. M. Cohen
THE work of Ahlie and his students and colleagues has brought
to the fore great amounts of descriptive information about
the early life history (ELH) stages of fishes gathered over many
years. These data are of broad provenance, many being the
results of original research by the Ahlstrom school, others being
taken from the literature. Only a scientist with Ahlie's capabil-
ities—an extensive knowledge of fishes and their ontogeny, a
fine sense of order in nature, and a critical intellect— could per-
ceive pattern in the bewildering diversity represented by the
early life history stages of fishes. As would any good scientist,
Ahlie questioned the meaning of these patterns, and it is chiefly
to further this inquiry that this symposium was convened.
Most students of comparative fish ontogeny know more about
adult fishes than ichthyologists who study adults know about
larval fishes; they have to. Ahlie stated in his lectures. "Larval
taxonomy is just an adjunct to adult taxonomy and you have
to start with the adults to know the larvae." Early on he dis-
covered that data from early life history studies did not always
confirm classifications based on adults alone. We all want to
know which data sets most closely approximate phylogenetic
relationships; how apparent conflicts best can be resolved; how
the data of ontogeny can be integrated into the overall field of
fish systematics? Answering these questions is not easy, espe-
cially within the framework dictated by the widespread adoption
of new methodologies in systematics, which claim to require
more stringent evaluation of characters than has been heretofore
customary. Many traditional character suites are being rejected
for purposes of elucidating phylogenies, and new data are needed
for testing. Our purposes m this volume are to state the bases
for what has come to be called larval fish taxonomy and to
consider the systematics of various groups of fishes in terms of
the rich and virtually untapped store of data offered by the study
of early life history stages.
My own objectives in the present paper are several. First of
all. I want to indicate the reasons, some obvious, some not, for
the nearly exclusive use of adult fishes in systematics, which has
prevailed until very recently. Secondly, I will briefly discuss the
conceptual and methodological framework of classification
within which early life history data is being used. Finally, I will
comment on the possible importance of early life history data
for the study of phylogeny with special reference to fishes.
Why Has There Been So Little Use of
ELH Stages in Fish Systematics?
The fact that most fish classifications are based entirely or
chiefly on the structure of adults was a source of concern to
Ahlie and remains so to many of us, although this Symposium
is an indication of positive change. I discuss below what may
be some of the reasons for a long preoccupation with adults.
In the first place, zoologists have been studying adults for a
longer period of time than they have early life history stages.
Although the dim beginnings of classification are often placed
with Aristotle, it was the great naturalists Aldrovandi. Belon.
Gesner. and Rondelet who in their cataloging of nature provided
our earliest adult fish classifications. Several technological de-
siderata would have prevented the study of early life history
stages during the 1 6th century when these early scientists were
at work. Even though lenses had been known for a long time,
appropriate microscopes were not invented until the 1 7th and
18th centuries (Singer, 1959) when another requisite advance
occurred, the use of alcohol and other fluids as a preservative
for zoological specimens (Singer, 1950). Techniques for clearing
flesh and staining bone and cartilage are modem acquisitions,
as is the use of x-ray photographs (Ahlstrom and Moser. 1981).
The invention of fine-mesh towing nets did not occur until 1 846
(Sverdrup. Johnson, and Fleming, 1942), deferring until rela-
tively recent times the availability of suitable collections of early
life history stages for scientific study.
The rearing of early stages is another valuable component of
the study of larval fish taxonomy, and although fish culture is
an ancient art, the staging of fry and their preservation and
microscopic study is technology-dependent and relatively re-
cent.
Lack of information on metamorphosis or of congruence of
larval and adult stages has also delayed the adoption of early
life history stages information into classification schemes. Of
course not many kinds of fishes demonstrate an ontogenetic
change as sudden and dramatic as do the eels, but the fact that
this particular transformation was not described until 1897
(Grassi and Calandruccio) indicates the long advance start held
by the use of adult stages. Even more recent have been discovery
of the Anoplogaster-Caulolepis relationship (Grey, 1955a), the
Gibberichthys-Kasidoron relationship (de Sylva and Eschmeyer,
1977), the Giganturidae-Rosauridae relationship (Johnson, this
volume), and the as-yet-unpublished identity of larval forms
such as Svetovidovia. These and other examples are described
in this volume. And indeed, even when the study of the devel-
opmental biology of vertebrates commenced, early emphasis in
the mid- 18th century was on classical embryology, the describ-
ing of processes and structures rather than on comparing them
(Rostand, 1964). Not until the early years of the present century
when fishery scientists began to use larval fishes in their inves-
tigations of commercial species and required identifications were
serious efforts made to compare data (Ahlstrom and Moser,
1981).
Until Ahlie commenced his now famous courses on larval
fishes, there were few places where a student could learn about
them; hence, there are only rare instances of attention being
paid to any potential value they might have in solving problems
in systematics. By now, in contrast, there are courses and sem-
inars available in a number of universities on the study of ELH
stages of fishes.
Another phenomenon that I believe has inhibited the use of
early life history stages in fish systematics is what I call the
curatorial mind set. Many curators of adult fish collections are
wary of microscopic specimens stored in vials. Although these
collections occupy small space, their maintenance and docu-
mentation are labor-intensive and their use is foreign to most
ichthyologists. There are many excellent collections of larval
fishes, but they are mostly in fishery, environmental and marine
biology laboratories— organizations that have no institutional
commitment to long-term collection storage. Collections that
document important publications or have potential value in
systematics should ultimately be deposited in a museum that
8
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
has a mandate to guarantee long-term archival storage and easy
access. Several such institutions that presently house larval fish-
es or are willing to do so are the Zoological Museum of the
University of Copenhagen, which maintains the extensive
worldwide collections taken during the Dana Expeditions, as
well as ones documenting the earlier classical studies on larval
fishes by Johannes Schmidt and his students, the Harvard Mu-
seum of Comparative Zoology, the Smithsonian Institution, and
the Natural History Museum of Los Angeles County. If collec-
tions of ELH stages are to realize their full potential in system-
atics, then it is timely for ichthyoplankton specialists to offer
good developmental series, especially illustrated ones, and for
museum curators to accept them.
Fossils have been studied for clues to the major classification
of fishes since the days of Louis Agassiz (Patterson, 1981a) and
to the extent that they were available have been widely consid-
ered as important adjuncts or indeed prerequisites to compre-
hending the phylogeny of particular groups. Although this view
is now receiving heavy criticism (Patterson, 1981b), the fact
remains that it did exist for many years and may have detracted
from the potential contribution of the non-fossil suites of char-
acters carried by early life history stages. Even so, students of
fossils and of larvae share a preoccupation with the caudal fin
skeleton, a structure that is often well preserved in fossils and
can be studied in two dimensions and which, during the course
of ontogeny, exposes a wealth of information of great value to
the systematist.
Because adult stages have been the chief source of characters
used in fish systematics, a perception has arisen that these char-
acters are in some way more useful or more indicative of a
phylogenetic classification than are the characters of early life
history stages. How did such a view arise? For many years,
systematists tended to concentrate on the search for conserva-
tive, "non-adaptive" characters (labeled the Darwin Principle
by Mayr, 1969). They discarded not only ones that they believed
were directly affected by the environment but also ones that
appeared to smack of convergence. It seemed reasonable and
proper, for example, to group together for phylogenetic purposes
fishes with one spine and five soft rays in the pelvic fin because
the character was apparently conservative, non-adaptive, and
non-convergent. On the other hand, it seemed wrong to group
together all fishes with canine teeth because the character was
apparently non-conservative, adaptive, and surely convergent.
With regard to larval fishes, Moser (1981) recently discussed
the occurrence of a large number of apparently highly adaptive
larval characters distributed across a broad taxonomic spec-
trum. He states, "Marine teleost larvae have evolved an enor-
mous array of morphological specializations, such that it seems
to me we are looking at a distinct evolutionary domain quite
separate from that of the adults. It is reasonable to assume that
these remarkable structural specializations are adaptive and re-
flect each species' solution to the challenge of survival in a
complex and demanding environment." My point here is that
if a systematist rejected adaptive characters (and many did),
then he would have been unlikely to use ELH stages, and this
may be another reason why they have not received sufficient
attention.
How Systematists Do Their Work
Even if systematists agreed among themselves about their
immediate goals and how best to achieve them, the task of this
Symposium would be daunting. But contemporary systematists
do not agree on either objectives or methodology. The concepts
that purport to link systematics to phylogeny are being actively
reassessed, and it is within the context of rapidly changing ideas
in systematics that our presentations and discussions will occur.
There are basically three conceptual methods now being used
by systematists, and although the bare bones of these methods
are easily comprehended, in practice they become more complex
and their independence from each other less clear. The interested
reader who is as yet unaware of the intense debate both between
and within the several schools of systematic classification is
referred to the pages of the journal Syslonatic Zoology for many
articles and references as well as ones cited in this section. A
recent description and comparison of the three methods is given
by Mayr (1981), who lists many important references. Although
1 do not propose to use very much space here on a redundant
treatment, 1 will briefly describe each method and comment on
its strengths and weaknesses.
The theoretically simplest method (or methods— there is more
than one algorithm, and there is disagreement on which is best)
is called phenetics or numerical taxonomy and is described in
detail by Sokal and Sneath (1963) and Sneath and Sokal (1973).
It is based on overall similarity. Many unweighted characters
are used to generate clusters of OTUs (operational taxonomic
units), which may be anything from individuals, populations,
or species to orders, classes, or phyla. The hierarchically ar-
ranged clusters, which lack a time dimension, are called phe-
nograms. Neither homology nor the fossil record are considered
in selecting characters. Each member of a cluster bears a closer
resemblance, although not necessarily genealogical relationship,
to other members of its cluster than it does to members of other
clusters. Some pheneticists claim that if a sufficient number of
characters is analyzed, any influence of convergence becomes
dampened and the phenogram will express phylogenetic rela-
tionships. Unfortunately, there seems to be no good way to
ascertain how many characters are needed. Other pheneticists
do not ascribe phylogenetic significance to their clusters and
merely claim to be representing overall similarity. Replicability
of results is the chief objective. Many classifications that purport
to be based on the methods of cladistics or evolutionary clas-
sification, upon close scrutiny appear to be basically phenetic.
There are apparently few fish classifications using ELH char-
acters, which are explicitly based on phenetic methods. One
example is a paper on Northeast Pacific cottid genera (Rich-
ardson, 1981a) which, according to the author, was not entirely
satisfactory for phyletic purposes. Ichthyologists who restrict
their data sources for a phenetic analysis to a single life history
stage should consider a study by Michener (1977), who gener-
ated four different phenetic classifications of a group of bees
based on different life history stages or character suites.
A second method is called cladistics or phylogenetic system-
atics, and although it has been more or less on the scene for
many years, it is only since the revision and translation into
English of its original presentation (Hennig, 1950, 1966) that it
has gained wide currency and is now used, either explicitly or
implicitly, by many systematic ichthyologists all around the
world but particularly in North America and western Europe.
A recent guide to the method is a book by Wiley (1981), and
the reader is advised to consult also Brundin ( 1 966) for a notably
lucid interpretation. Cladistics requires a stringent evaluation
of characters. Primitive or generalized ones (called plesiomor-
COHEN: ONTOGENY, SYSTEMATICS, PHYLOGENY
phic) for the group being analyzed are discarded for purposes
of generating a phylogenetic classification; only derived char-
acters (apomorphic) are of value, and monophyletic groups are
defined by the degree to which they share such characters (syn-
apomorphy). The distribution of derived character states among
a monophyletic assemblage of taxa is analyzed and used to
generate an hierarchically arranged chart called a cladogram, in
which each node or branching point on the diagram gives rise
to two branches that are interpreted as genealogical lineages and
are called sister groups. In instances in which the data do not
allow the unambiguous definition of two branches, more are
often used. Each member of a monophyletic group is more
closely related genealogically to other members of its group than
it is to members of other groups. More than one cladogram can
be generated with the same data set, and the most parsimonious,
that is, the one requiring the fewest evolutionary steps, is taken
as the most natural or best. According to Panchen ( 1 982), prob-
lems in logic invalidate the use of parsimony in cladistics. Not
all cladists agree about precisely what a cladogram represents,
but some interpret it directly as a phylogenetic classification.
One of the greatest problems in using cladistics is the difficulty
in evaluating character states for primitiveness or degree of
derivation. Two methods have been used; one involves onto-
genetic stages and will be discussed later in this paper. A second
method, called out-group comparison (Wiley, 1981, gives a good
description), is the most subjective part of the entire cladistic
procedure and to a certain degree may involve circular reason-
ing. A practical problem that cladistics has not yet conquered
is that of naming, for classifications must be used by many who
have no interest in theory, and naming categories on a strictly
genealogical basis raises many problems, as does the practice
followed by some cladists of naming all branching points. Some
attributes of ELH stages that might be considered unsuitable
for use in evolutionary classification are available for use in
cladistics. One example concerns character stages that are in-
terpreted as being highly adaptive rather than conservative. If
polarity can be ascertained, then so-called adaptive characters
are available. Rates and sequences of ontogenetic change also
constitute potentially valuable character suites.
The third method, presently called evolutionary classification,
is more difficult to define and discuss. It has a long history and
an extensive literature (Mayr, 1981). The methods of evolu-
tionary classification are eclectic and generally more subjective
than those of phenetics and cladistics. They do not easily lend
themselves to overall generalization. Characters are selected and
weighted by paying particular attention to homology and con-
vergence; to the extent that they are available, evidence from
embryology and palaeontology are also used. Primitive char-
acters are admitted to the system. Data are used from ecolog-
ically oriented facets of evolution such as selection, competition,
predation, and ecological biogeography. Historical biogeogra-
phy, rate of evolution, and genetics are also considered. An
hierarchical classification is derived, which has an inferred time
axis and which may generally reflect genealogical relationships.
However, degree of phenetic difference in selected characters,
which is interpreted as reflecting degree of genetic difference,
may be considered along with branching pattern in converting
a strict genealogy into a classification. Patterson (1981b) has
discussed and criticized such procedure. Whatever may be phy-
letic relationships, the definition of taxa is essentially subjective,
and each member of a group is not necessarily more closely
related genealogically to other members of its group than it is
to members of a different group. The test for goodness of a
classification is pragmatic; if it has high predictive value it is
good. (By prediction is meant the degree to which a classification
encompasses additional data.) In commenting on evolutionary
systematics Panchen (1982) writes that it, "has always been
somewhat ad hoc in its procedure, yielding good results with
competent taxonomists and bad with incompetent ones. The
standard warks [sic] on procedure . . . are to some extent ra-
tionalizations of a tradition that is too largely intuitive."
As a summary, I have tried to compare in Table 1 some of
the techniques, objectives, and assumptions of the three meth-
ods. Phenetics requires the fewest assumptions but would seem
to offer the systematist a classification with the least information
value. Cladistics has the most constraints, so many and so strin-
gent in fact, that they may limit its practical use, although the
method is particularly valuable in indicating areas for which
additional or more suitable data are required. Misuse of cla-
distics may soon rival the long-time abuse by systematists of
parametric statistics. Evolutionary classification tries to include
the most information from the most sources, but the methods
for doing so are not very well formalized. Cladists treat their
method of classification as a general theory of biology (Nelson
and Platnick, 1981), a forcing function among all evolutionary
phenomena, which must therefore comply with a parsimonious
model derived entirely from character state analysis. Evolu-
tionary classification, on the other hand, incorporates infor-
mation from a wide variety of biological phenomena and to
that extent is forced, rather than forcing. Predictability, as a test
of goodness for a classification, is more pragmatic and logically
less satisfying than is parsimony. Perhaps an important question
for theoretical systematists to consider is the formulation of
comparable definitions for replicability, parsimony, and pre-
dictability.
Ontogeny and Fish Phylogeny
Louis Agassiz, who fought the idea of organic evolution, pro-
posed a "threefold parallelism" of arranging organisms in a
series or classification. His three parallels were palaeontology,
what we would now consider to be homology, and ontogeny.
Even though he failed to interpret the parallels as evidence for
evolution, his keen perception of the fact that they do exist in
nature and are somehow interrelated has elicited extensive com-
ment and reinterpretation (see especially Gould, 1977) and is a
suitable point of departure for addressing the importance of
ontogeny as a source of information about homology, the bio-
genetic law, developmental stages as alternatives to outgroup
comparisons in cladistics, paedomorphosis, and the application
of life history stages to phylogenetic inquiry.
If characters are the meat and muscle of classification, then
homology surely shapes the skeleton on which phylogenetic clas-
sifications are arranged. The worth of any allegedly phylogenetic
classification is no better than the degree to which homology
has been assessed, and how to do this is a major problem for
the systematist. Like the weather, everyone talks about homol-
ogy but does nothing about it— or almost nothing. The concept,
which is so pervasive in the study of phylogeny and in evolution,
has been with us since pre-Darwinian times, although not always
in the way that we understand it today. The great comparative
anatomist Owen defined it in 1866 as follows; "A 'homologue'
10
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 . Comparison of Three Methods Used in Biological Classification.
Evolutionary'
Character weighting
Convergence
Homology
Fossil History
Eco-evolutionary Data
Rale of Evolution
No. of Characters
No. of Specimens
Branches from a Node
End Product
Test of Goodness
No
Not Considered
Not Considered
Not Considered
Not Considered
Not Considered
Many
Few
Two to Many
Perhaps a Phylogeny
Replicability
Yes
Important
Important
Not Important
Not Important
Not Important
One to Medium
Few to Many
Two when Possible
Phyiogenetic Classification
Based on Genealogy
Parsimony
Yes
Important
Important
Important
Important
Important
One to Medium
Few to Many
Two to Many
Phyiogenetic Classification Based
on Genealogy and Degree of
Difference
Predictability
is the same pail or organ in different animals under every variety
of form and function." He goes on to note, however, that some
earlier workers defined the concept as we now define analogy.
But our problem remains identical with that of Owen— how to
define same. In a recent discussion of homology, Patterson (1982)
proposed similarity in ontogeny as part of a test of homology.
But the use of similarity in development to help define Owen's
"same" is tautological.
Palaeontologists proceed in a basically circular fashion in their
use of homology. They depend upon a time series to trace the
history of transformed states of a presumably homologous char-
acter along a sequence that is interpreted as a genealogy. But of
course the characters are considered homologous because they
are part of a genealogy. Whether they admit to it or not, most
systematists use pure phenetics in the search for homology, and
although this common sense, intuitive, non-scientific approach
works much of the time, still, many systematists have misin-
terpreted as homologues characters that are actually analogous
and have filled the literature with many misdiagnosed conver-
gences. In comparative vertebrate anatomy and systematics, the
convention has grown up that certain organ systems are more
conservative than others and therefore provide a better method
for detecting homologies. The nervous system is generally con-
sidered the best, the skeleton the next best, followed by viscera
and muscles, with the integument the least good. In fishes, for
example, Freihofer (1963, 1970) has used the patterns of the
ramus lateralis accessorius and ramus canalis lateralis nerve
systems relative to elements of the skeleton to propose groupings
of fishes. But even here the possibility of convergence cannot
be ignored (Gosline, 1968), and again the problem of circularity
arises because many ichthyologists define osteological features
on the basis of their topographic relation to elements of the
nervous system. Another example relates to homologies of pho-
tophore series in lantemfishes as determined by studies of their
innervation (Ray, 1950). Here also, the conclusions based on
this method appear to be equivocal (Moser and Ahlstrom, 1 972).
A direct method for demonstrating the homology of structures
would be to trace them back during development to their an-
lagen. De Beer (1951) has commented on the apparent failure
of experimental embryology to validate this approach. Even so,
a survey of the development of bony structure during fish on-
togeny presented by Dunn ( 1 983b) lists some observed instances
of losses, gains, and modifications, chiefly in the caudal fin skel-
eton, which interpret homologies in adult structure; unfortu-
nately, these instances are too few. Ahlie had a long interest in
the caudal fin skeleton, particularly of flatfishes, and the com-
pletion of his work by colleagues hopefully will constitute an
additional contribution to the use offish ontogeny in identifying
homologous structures.
The concepts of ontogeny and homology are intimately as-
sociated in the idea that the study of early life history stages of
an organism will reveal its adult ancestral stages— ontogeny re-
capitulates phylogeny— as proposed by Ernst Haeckel in the
latter half of the 19th century. Taken at its most extreme, the
biogenetic law has been interpreted as meaning that an entire
genealogy is encapsulated in an ontogenetic series. If adults of
extant species of a group were to be matched up with their closest
approximations in an ontogenetic series, homology would un-
fold before our eyes. Of course its value to us in unraveling
phylogeny would be redundant, because phylogeny would be
there as well. It was soon evident however that the biogenetic
model is far too crude to approximate nature. The embryologist
von Baer had previously formulated four "laws" or general
propositions about embryology that have been restated in var-
ious forms by many authors and applied to the interpretation
of phylogeny. The following are taken from De Beer ( 1 95 1 ): ( I )
In development from the egg the general characters appear be-
fore the special characters. (2) From the more general characters
the less general and finally the special characters are developed.
(3) During its development, an animal departs more and more
from the form of other animals. (4) The young stages in the
development of an animal are not like the adult stages of other
animals lower down on the scale, but are like the young stages
of those animals. These propositions are useful generalizations
and we can all think of obvious instances of fish , ontogeny that
can be interpreted by one or more of them . Consider for example
the bilaterally symmetrical larvae of flatfishes, the early presence
and subsequent loss of a swimbladder in stromateoids (Horn,
1970a), the sequence of fusions during ontogeny in the caudal
fin skeleton of myctophids (Ahlstrom and Moser, 1976), the
ontogeny of the upper jaw bones and dentition in notosudids
(Berry, 1964a), and the presence of a pectoral fin in larval Tac-
tosloma and its loss in adults (Ahlstrom, lecture notes). On the
other hand, a plethora of early life history stages of fishes man-
ifests character states that represent morphological specializa-
tions occurring early in development. Consider the egg stages
of macrourids with their hexagonal patterns, atherinomorphs
with their filaments, and argentinoids with their pustules. Other
COHEN: ONTOGENY, SYSTEMATICS, PHYLOGENY
11
instances for which it is difficult to accept that ontogeny has
recapitulated phylogeny include the leptocephalus of eels, the
stalked eyes of assorted larval bathylagids, myctophids and Idi-
acanthus. the elongated guts of larval melanostomiatids, the
extensive armature of many spiny-rayed fishes during their lar-
val stages, and the produced fin rays found in many kinds of
larval fishes. Examples of all of these are illustrated and de-
scribed in this volume. With regard to proposition three in
particular, Ahlie often pointed out instances of fishes that were
easily distinguished as larvae but became more similar in ap-
pearance as adults; one example is Bathylagiis milleh and B.
pacificus; Myctophum aurolaternalum and other myctophid
species is another. Von Baer's propositions as applied to phy-
logeny are tidy and appealing but are completely operative only
under the rather special condition that major evolutionary
changes (except for paedomorphosis) are restricted to the adult
stage (Gould, 1977; Fink. 1982).
For cladistic analysis, the polarization of characters through
direct observation of their transformation during ontogeny has
been discussed by Nelson (1978) and others as an alternative
to the often unsatisfactory indirect method of outgroup com-
parison. Such use of ontogeny, which depends on von Baer's
first three propositions, has been analyzed by Henning (1966),
who noted its uncertainty. As examples from fish ontogeny given
above indicate, ontogeny could replace or corroborate outgroup
comparison but only to the extent that the biogenetic law is
valid for a particular situation. Patterson's (1982) statement,
"that ontogeny is the decisive criterion in determining polarity,"
would seem to be based on limited acquaintance with ELH
stages.
Paedomorphosis refers to the presence in adults of larval char-
acters (De Beer, 1951) and has been variously considered as
insignificant to very important in evolution. For fishes at least,
I think the latter is the case. As one example, small adult size
could be considered a particularly widely distributed neotenic
character. In his discussion of paedomorphosis and cladistics.
Fink (1982) remarked that it is difficult to identify this phe-
nomenon without paired taxa, but surely this is not always true.
Although the relationships of the curious little fish Schindleria
are unknown, it would be difficult to deny that it has many
neotenic characters (Watson, Stevens and Matarese, this vol-
ume). On a larger scale paedomorphosis may have been im-
portant in establishing novel phyletic lines as well as isolated
species or genera, and the study of ELH stages will be essential
in detecting these divergences.
I end this essay by noting that the most important use of all
for information about fish ontogeny may be providing characters
for charting fish phylogeny rather than theories about phylogeny.
Distinguishing and identifying species for purposes of fish bi-
ology and management has been the chief use for what is called
larval fish taxonomy, and the large resulting literature is sum-
marized in this volume. Many of the same descriptive data are
of apparent value for purposes of grouping similar species or
other taxa for phyletic purposes. Published examples of syn-
thesis are far fewer than of descriptions, but accounts using each
of the three methodologies previously described are available,
either cited in this volume or presented here as original research.
ELH characters can meet many methodological constraints and
will be used increasingly by ichthyologists. To what advantage
remains to be seen, but the prognosis is good.
Life Sciences Division, Los Angeles County Museum of
Natural History, 900 Exposition Boulevard, Los
Angeles, California 90007.
Early Life History Stages of Fishes and Their Characters
A. W. Kendall, Jr., E. H. Ahlstrom and H. G. Moser
Patterns of Teleost Early
Life History
IN discovering that Atlantic cod lay free-floating planktonic
eggs which develop into pelagic larvae, G. O. Sars, in 1865
(see Hempel, 1979; Ahlstrom and Moser, 1981) had also come
upon an example of the widespread life history pattern of marine
fishes. Most marine fishes, regardless of systematic affinities,
demersal or pelagic habits, coastal or oceanic distribution, trop-
ical or boreal ranges, spawn pelagic eggs that are fertilized ex-
ternally and float individually near the surface of the sea (Fig.
5). These eggs range from about 0.6 to 4.0 mm in diameter
(mode about 1 mm) and generally are spherical. Within a species
there is little variation in egg characters such as size, number
and size of oil globules, and pigmentation and morphology of
the developing embryo. Development time is highly tempera-
ture dependent and also species-specific. The eggs hatch into
relatively undeveloped yolk-sac larvae which swim feebly and
rely on their yolk for nourishment while their sensory, circu-
latory, muscular, and digestive systems develop to the point
that they can feed on plankton. Even these yolk-sac larvae have
characters (pigment patterns, body size and shape, myomere
number) that reflect their heritage. After the yolk is utilized,
they develop transient "larval" characters such as pigment pat-
terns and, in some, specialized head spines and fin structures
that are apparently adaptive for this phase of their life history.
During this period more characteristics of the adult (e.g., me-
ristic characters) gradually develop. At the end of the larval
stage, they may go through an abrupt transformation to the
juvenile stage, particularly if they move from a pelagic to de-
mersal habitat, or the transformation may be gradual. In some
fishes, there is a prolonged and specialized stage between the
larval and juvenile stages. These pelagic (often neustonic) forms
eventually transform into demersal juveniles. The juvenile stage
is characterized by specimens having the appearance of small
adults— all fin rays and scales are formed, the skeleton is almost
EGGS
YOLK
SAC
PRE
FLEXION
FLEXION
POST
FLEXION
>
<
>
m
JUVENILE
Fig. 5. Early life history stages of Trachurus symmelricus from Ahlstrom and Ball (1954),
KENDALL ET AL.: ELH STAGES AND CHARACTERS
13
END POINT EVENTS
TERMINOLOGY
Primary developmental
stages
Transitional stages
Subdivisions
OTHER
TERMINOLOGIES
Hubbs, 1943,1958
Sette, 1943
Nikolsky, 1963
Hattori, 1970
Balon, 1975 (phases)
Snyder, 1976,1981 (phases) \
E q q
Larva
Juvenile
1 ' '
1
Yolk sac
larva
Transforma
lion larva
Early
Middle
Late
Piftlexion
Flexion
PnMflexinn
larva
larva
larva
Pelagic or
special juven
'
1
E m b
y o
Proiarva
Post 1 a rv a
Prejuvenile
'
1 1
Larva
Post larva
Embryo
Prelarva
Cleavage egg
Embryo
Eleulhero
embryo
Protoptery-
qiolarva
Pterygiolarva
1
1
1
1
1 1 1 1
Protolarva
M e s 0-
larva
#
M e t a 1 a
V a
Fig. 6. Terminology of early life history stages.
completely ossified, the larval pigment pattern is overgrown or
lost and replaced by dermal pigment similar to that of the adults,
and the body shape approximates that of the adults.
Although this is the most frequently observed life history
pattern, there are many variations (see Breder and Rosen, 1 966)
often related to increased parental investment in individual
progeny with a concomitant decrease in fecundity and larval
specializations. There is scant information on the young of many
deep-sea fishes, and this may be due in part to life history
strategies that do not include eggs and larvae that occur in the
epipelagic zone (where most of the collecting is done). Marshall
(1953) discussed life history adaptations of these fish such as
the production of few, large yolky eggs that hatch into relatively
advanced larvae. These young may remain far below the more
productive surface layers, and thus not be susceptible to most
sampling procedures. Markle and Wenner (1979) cite evidence
for demersal spawning of two species of groups (Alepocephal-
idae, Zoarcidae) that are seldom collected in the plankton as
larvae.
Many coastal marine and nearly all freshwater fishes lay de-
mersal eggs which are generally larger than the I mm mode of
pelagic eggs. In such fish development from hatching through
juvenile stage is direct and the larvae gradually attain adult
characters of shape, pigmentation, and meristic features. The
demersal eggs frequently are adhesive and laid in some sort of
nest. Parental care of the nest is observed in many species, and
this care may extend to the larvae after hatching (e.g., mouth
brooding in cichlids, ariids). Parental care takes another form
in Sehastes. where development through the yolk-sac stage takes
place in the ovary and first-feeding larvae are extruded. Vivi-
parity, in which nourishment is supplied by maternal structures,
has evolved many times (e.g., poeciliids, some zoarcids, em-
biotocids), whereby the larval stage is bypassed and the fish are
extruded ("bom") as juveniles (Wourms, 1981).
Early Life History Stages
Between spawning and recruitment into the adult population,
most fishes undergo dramatic changes in morphology and hab-
Table 2. Examples of Characters of Pelagic Eggs that May Be
Useful for Systematic Studies of Certain Fishes.
Character slates
Systematic groups
Egg size
Egg shape
Envelope
sculpturing
Oil globule
position
Embryonic
characters
< 1 mm->5 mm
>3 mm->5 mm
Round — oblong
Varying distances between
pores
Varying length/density of
filaments
Anterior to posterior in
yolk sac
Slate of development of
various organs/organ sys-
tems at various develop-
mental mileposts
Pleuronectidae
Anguilliformes
Engraulidae
Ostraciontidae
Gadidae
Atheriniformes
(Exocoetidae)
Perciformes
Gadidae
14
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 7. Examples of features of yolk-sac larvae of teleosts. (A-C). Paracallionymus costatus. A. soon after hatching 0.98 mm NL; B. 1.8 mm
NL; C. 1.9 mm NL. From Brownell (1979). Features demonstrated in; (A) include the small size of the larva, the lack of an oil globule, the
segmented yolk, and the dorsally arranged melanophores; (B) demonstrates the migration of melanophores ventrally and the formation of the
anus producing a preanal finfold; (C) demonstrates further ventral migration of melanophores, beginning of larval pectoral fin formation, the
decrease in yolk-sac size, and beginning of pigment in the eye; (D) Diplodus sargus. 2.4 mm NL. From Brownell (1979). Single pigmented oil
globule posterior in the unsegmented yolk and a short preanal finfold are demonstrated; (E) Trachurus I. capensis. 2.2 mm NL. From Brownell
(1979). Single pigmented oil globule anterior in segmented yolk with moderately long preanal finfold demonstrated; (F) Cololabis saira. 5.1 mm
SL. (original). Well-developed, heavily pigmented yolk-sac larva at hatching with notochord flexion beginning and some caudal rays formed; (G)
Argentina silus. 1.1 mm. Redrawn from Schmidt (1906c). A large but poorly developed yolk-sac larva at hatching with a large oil globule; and
(H) Hippoglossus slenolepis. 9.5 mm. From Pertseva-Ostroumova (1961). A large but poorly developed yolk-sac larva at hatching with no oil
globule.
its. As mentioned earlier, at hatching, particularly in marine
fishes with pelagic eggs, the fish is in an extremely undeveloped
state and then, as a free-living individual, it gradually develops
the adult characters. This process is continuous, but there are
morphological and ecological mileposts that are significant in
the life of the fish and which allow us to subdivide this process
so that we can communicate results of our studies and compare
different fishes at the same moment in development.
Fish early life history has been and continues to be studied
from a number of different perspectives (Ahlstrom and Moser,
1976). Some studies deal directly with embryology and later
ontogeny, others emphasize functional morphology of larval
structures, apply larval features to taxonomic and systematic
studies, investigate the ecology of eggs and larvae, or use these
stages to address fishery-related problems such as assessment
of spawning stock size and recruitment success. All of these
studies have in common the need to subdivide early life history
and communicate information based on processes and events
occurring during these subdivisions. As with any communica-
tion, it is vitally important to use terms that are clearly defined
and this is particularly true with the diverse disciplines that are
involved in larval fish studies. Historically, several disciplines
have used different names for the same stage, or subdivided
development differently [see Okiyama (1979a) and Fig. 6 in this
paper]. This has led to confusion rather than communication.
Several criteria seem appropriate for defining stages of de-
velopment to be used by students of any discipline. The variety
of developmental patterns should be recognized and the defi-
nitions should apply to as many patterns as possible. Thus,
stages should be based on very widespread, fundamental fea-
tures of development. The stages should have some significance
in the life history of the fish, both morphologically and func-
KENDALL ET AL.: ELH STAGES AND CHARACTERS
15
From demersal eggs
From pelagic eggs
Clupea harengus harengus
egg diameter = 1.2-1. 5mm
NL at hatclning = 4.9mm
Etrumeus teres
egg diameter = 1.3mm
NL at hatching = 4.8mm
Krevanoski 1956
Mito 1961
O
Mukhacheva and Zviagina 1960
Gadus macrocephalus
egg diameter = 0.8-1. 4mm
NL at hatching = 3.6mm
Colton and IWarak 1961
Gadus morhua
egg diameter = 1.1 -1.9mm
NL at hatching = 3.6mm
Lepidopsetta bilineata
egg diameter = 1.02-1. 09mm
NL at hatching = 3.9mm
Isopsetta isolepis
egg diameter = 0.90-0. 99mm
NL at hatching = 2.9mm
Pertseva-Ostroumova 1961
Richardson et al 1980
Fig. 8. Newly hatched yolk-sac larvae of related fishes with pelagic and demersal eggs of comparable sizes.
tionally, such as a particular type of nourishment or locomotion.
Also the endpoints for the stages should be easily observed and
sharply defined.
The most general scheme of terminology of early development
of fishes includes (Fig. 5):
The "egg stage" (spawning to hatching). The egg stage is used
in preference to the embryonic stage because there are characters
present during this stage other than just embryonic characters
(e.g., those associated with the egg envelope).
The "larval stage" (hatching to attainment of complete fin
ray counts and beginning of squamation). One of the funda-
mental events in development of most fishes is the flexion of
the notochord that accompanies the hypochordal development
of the homocercal caudal fin. It is convenient to divide the larval
stage on the basis of this feature into "preflexion." "flexion,"
and "postflexion" stages. The flexion stage in many fishes is
accompanied by rapid development of fin rays, change in body
shape, change in locomotive ability, and feeding techniques.
The "juvenile stage" (completion of fin ray counts and be-
ginning of squamation until fish enters adult population or at-
tains sexual maturity).
Transitional stages can also be recognized: the "yolk-sac larval
stage" (between hatching and yolk-sac absorption); and the
"transformation stage" (between larva and juvenile). Meta-
morphosis occurs during this stage and is considered complete
when the fish assumes the general features of the juvenile.
The life histories of some fishes include other specialized
ontogenetic stages that have received various names. In some
cases, these are the generic names under which these stages were
described before they were recognized as larvae of other species
(e.g., the leptocephalus stage of Anguilliformes, the scutatus
stage of Anlennarius. the vexillifer stage of Carapidae. and the
kasidoron stage of Gihhertchthys). In other cases, consistent fea-
tures of development of a group permit useful subdivisions of
stages (e.g.. in leptocephali the engyodontic and euryodontic
stages).
The Egg Stage
Hempel (1979) reviewed the egg stage relative to fisheries
investigations. Ahlstrom and Moser (1980) presented a concise
review of the range of characters observed in pelagic fish eggs,
particularly those useful in identifying eggs in plankton samples.
Sandknop and Matarese in this volume also discuss this subject
in detail. The characters that have proven useful for egg iden-
tification include egg size and shape, size of perivitelline space,
yolk diameter and character (homogeneous or segmented), num-
ber and size of oil globules, texture of the egg envelope (smooth
or with protrusions), pigment on the yolk and embryo, and
characters of the developing embryo (relative rate of develop-
ment of various parts, body shape, number of somites) (Table
2).
The egg stage has been subdivided by a number of workers
(e.g., Apstein, 1909). Fishery biologists need to determine the
age of eggs at the time of collection for production, drift, and
mortality estimates. Embryologists have designated stages to
coincide with significant developmental features. While the stages
of fishery biologists are designed to divide the embryonic stage
into several easily recognized portions, embryologists are more
16
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 3. Examples of Use of Characters of Early Life History Stages in Taxonomic and Systematic Studies. X indicates range of stages
and taxonomic levels at which characters vary. (X) indicates infrequent state.
Developmental stage
Character
Lar\ac
Taxonomic level
Yolk-
sac
Pre-
flexion
Flexion
Post-
flexion
Trans-
forma-
tion
Rpfprenf**^
Spe-
cies
Genus
Family
Order
IX \. IK 1 \,lkVV«^
Egg
Keyed to Table 4
(X)
X
X
(X)
X
20
X
X
(X)
(X)
X
20,29
(X)
X
X
X
X
X
X
(X)
X
2,38
1, 2, 11, 19, 24, 27, 39
(X)
X
X
X
11, 19,24,27,39
X
X
X
X
X
1,2,3.5, 11. 15, 17, 19,
20,25.27.28.33,34
X
X
(X)
X
X
X
(X)
X
27,38
19
X
X
X
X
X
X
X
2, 3,4. 5. 10, 11, 13, 14,
19,20,23. 24,25, 26, 27,
28, 29, 31», 33, 37, 40
X
X
X
X
X
28,33,35,36,38
X
X
X
X
X
X
X
X
X
X
1,2,3,4,8,9, 11, 13, 14,
15, 17, 19, 20,21,22,
25,27,28,29,33,36,
38, 39, 40
X
X
X
X
X
X
X
9, 11,23,24,25,27,36.
38,40
X
X
X
X
X
1,9, 14,23,27,29,33
X
X
X
X
X
X
X
14, 27, 29
27
X
X
X
8. 10, 14
X
X
X
X
X
X
X
36
8, 14, 15.33
X
X
X
X
X
20, 33, 38. 39
(X)
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
14,20,29,33,38
8, 10. 14,20,33
14,20,33
10, 14,20
29
X
X
X
6, 19,20,30,32
(X)
X
X
X
(X)
X
X
X
X
X
X
X
X
X
X
7, 16, 19,23.29,33,40
11,27
12, 14,21
X
X
X
X
X
X
X
X
10, 11,22,23,29,30,39
(X)
X
X
X
X
13, 14,20,26,27,34
Meristic characters
Fin spines/soft rays
Principal caudal rays
Pelvic fin
Dorsal/anal fin
Pectoral fin
Vertebrae
Branchiostegals
Gill rakers
Larval characters
Body shape
Snout shape
Pigment patterns
Head spines
Fin ray elongation
Fin ray ornamentation
Fin ray serration
Pinfold size/shape
Preanal finfold
Pectoral size shape
Larval gut
Shape
Length
Larval eye
Shape
Stalked
Choroid tissue
Migration
Other characters
Egg characters
Osteological development
Scale formation
Photophore formation
Size at developmental stage
Fin development sequence
• Emphasis on oil globule placement in yolk-sac larvae.
interested in tracing the sequence of development. The em-
bryologist's approach will probably provide more useful infor-
mation for systematic investigations.
Although excellent, early descriptive work was done on teleost
embryology (e.g. Wilson, 1891), comparative research on de-
velopment needs to be done to allow an evaluation of its value
to syslematics, a subject that has proven so fruitful among in-
vertebrates. It appears, from the characters that have been stud-
ied in greatest detail, that convergence may overshadow phy-
letically significant information. For instance, the egg envelope
sculpturing on Pleuronichthys, a pleuronectiform, was found
even on scanning electron microscope examination to be quite
similar to that on Synodus, a myctophiform (Sumida et al.,
1979). Phylogenetically diverse fishes often have round pelagic
eggs, about 1 mm in diameter, with a single oil globule. Demersal
eggs from equally diverse fishes are generally larger than I mm
and develop a vitelline circulatory system. Yolk segmentation
seems to be a character of more primitive fishes, but some
carangids and other perciforms have yolks that are secondarily
segmented in an evolutionary sense. Detailed studies are needed
to sort out these and other features of the teleost egg and its
embryonic development in a systematic context.
KENDALL ET AL.: ELH STAGES AND CHARACTERS
17
Table 4.
Some Contributions in Which Ontogenetic Characters have been used to Examine Systematic Relationships (Updated from
Ahlstrom and Moser, 1981).
References
Dale
Ciroup dealt with
Egg
Stages
Ur-
vac
Juv
ad
Larval characters showing
relationships
No.
Among
spe-
cies
Among
genera
Among
subfam-
or Among
families orders
1,3.5
Ege. V.
1930,53,57
Paralepididae
—
+
+
X
X
2
Bertelsen. E.
1951
Ceratioidei
—
+
+
X
X
X
4
Bertelsen. E., and N. B. Marshall
1956
Minpinnati
—
+
+
X
X
X
6
Pertseva-Ostroumova. T. A.
1961
Pleuronectidae
+
+
+
X
X
7
Berry, F. H.
1964a
Mar. teleosts
—
+
—
X
g
Pertseva-Ostroumova, T. A.
1964
Myctophidae
—
+
—
X
9
Gutherz, E. J.
1970
Bothidae
—
+
—
X
10, 14
Moser. H. G., and E. H. Ahlstrom
1970, 74
Myctophidae
—
+
+
X
X
X
11
Mead. G. W.
1972
Bramidae
—
+
+
X
X
12
Ahlstrom, E. H.
1974
Stemoptychidae
—
+
+
X
13
Johnson. R. K..
1974b
Scopelarchidae
—
+
+
X
X
15
Okiyama. M.
1974a
Myctophiformes
—
+
—
X
X
16
Potthofr. T.
1974
Scombndae
—
+
+
X
17
Richards, W. J., and T. Potthoff
1974
Scombridae
—
+
+
X
18
Aboussouan. A.
1975
Carangidae
—
+
—
X
19
Ahlstrom, E. H.. J. L. Butler, and
B. Y. Sumida
1976
Stromateoidei
+
+
+
X
X
X
20
Ahlstrom. E. H.. and H. G. Moser
1976
Mar. teleosts
+
+
+
X
21
Ahlstrom. E. H., H. G. Moser, and
M. J. OToole
1976
Myctophidae
—
+
+
X
22
Bertelsen. E., G. Krefft, and N. B.
Marshall
1976
Notosudidae
—
+
±
X
X
23
Futch. C. R.
1977
Bothidae
—
+
—
X
X
24
Moser. H. G., E. H. Ahlstrom, and
E. Sandknop
1977
Scorpaemdae
—
+
±
X
X
X
25
Okiyama, M., and S. Ueyanagi
1978
Scombridae
—
+
—
X
X
26
Powlcs. H.. and B. W. Stender
1978
Sciaenidae
—
+
±
X
27
Kendall. A. W.. Jr.
1979
Serranidae
—
+
+
X
X
28
Ueyanagi, S., and M. Okiyama
1979
Scombridae,
Istiophoridae
—
+
+
X
29
Amaoka. K.
1979
Pleuronectiformes
(in part)
—
+
—
X
X
30
Dotsu. Y.
1979
Gobiidae
+
+
—
X
31
Suzuki. K.. and S. Hioki
1979a
Percoidei
+
+
—
X
X
32
Mito. S.
1979a. b
Mar. teleosts
+
—
—
X
X
33
Okiyama. M.
1979b
Myctophoidei
—
+
—
X
34
Potthoff. T.. W. J. Richards, and
S. Ueyanagi
1980
Scombrolabracidae
—
+
+
X
X
35
Zahuranec, B. J.
1980
Myctophidae
( Na nnobrach lu m)
—
+
+
X
X
36.37
Richardson. S. L.
1981a,c
Cottidae
—
+
+
X
38
Washington, B. B.
1981
Cottidae
—
+
X
X
39
Johnson. R. K.
1982
Scopelarchidae
Evermannellidae
—
+
+
X
X
X
40
Kendall, A. W., Jr., and B. Vinter
1984
Hexagrammidae
-
+
+
X
X
The Yolk-sac Larval Stage
At hatching, larvae can be at various states of developmenl,
dependent to a large degree on the size of the yolk (Fig. 7).
Larvae from eggs with small yolks are less developed at hatching
than those that hatch from eggs with larger yolks. Since the bulk
of maiine fish spawn eggs that are about I mm in diameter and
have a narrow perivitelline space, the yolk is only slightly less
than I mm. Larvae from such eggs generally lack a functional
mouth, eye pigment, and differentiated fins. They possess a large
yolk sac relative to the size of the lai~va which supplies nour-
ishment while the larvae develop to become self-feeding. Newly
hatched larvae from demersal eggs are generally further ad-
vanced in development than lai^ae from pelagic eggs of com-
parable size (Fig. 8). In these and other fish with large eggs,
hatching may be delayed until the yolk sac is absorbed and the
larvae are ready to feed at hatching, having bypassed the yolk-
sac larval stage. The delayed absorption of yolk reaches an ex-
treme in fishes such as salmonines in which the yolk-sac larva
transforms directly into a juvenile; Hubbs (1943) proposed the
term "alevin" be applied to this yolk-sac larval stage.
At hatching, locomotion and orientation of most yolk-sac
larvae are aided by a continuous median finfold (dorsal, caudal,
anal) and larval pectoral fins. During egg development, many
fish embryos develop melanophores that originate in the neural
18
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
,r ..'—v ^„.-«n.-)-'-T' ^
Fig. 9. Examples of teleost larvae illustrating extremes of some systematically useful larval characters. (A) Myctophum aurolaternatum. 26.0
mm (Moser and Ahlstrom, 1974). Note stalked oval eye with choroid tissue, trailing gut, and dorsal fin developing in finfold; (B) Epinephelus
sp.. 8.4 mm (Kendall, 1979). Note elongate, serrate dorsal and pelvic spines; (C) Adioryx (Holocentrus) vexillarius. 8.5 mm (McKenney, 1959).
Note head spines; and (D) Lopholatilus chamaeleonticeps, 6.0 mm (Fahay and Berrien, 1981). Note spines on head and body.
crest and are generally aligned along the dorsal surface of the
embryo. During the yolk-sac stage, these melanophores move
laterally and ventrally to establish the beginning of the larval
pigment pattern. Orton (1953a) describes these events in detail
in Sardinops sagax. This realignment may begin during the late
embryonic stages, before hatching. Some species hatch with few
if any melanophores, and when they first appear, they are in
ventral positions. Apparently, the pigment cells migrate before
pigment formation occurs.
The presence and position of oil globules in yolk-sac larvae
vary and can be of diagnostic value. In fishes with single oil
globules, it can be far forward (e.g., labrids, most carangids,
muUids, and lethrinids), in the middle of the yolk sac (e.g.. some
clupeids, serranids, and argentinids), or more usually near the
rear of the yolk sac. The shape and relative size of the yolk sac
itself are variable and provide additional taxonomic characters.
In summary, although the yolk-sac stage starts at hatching
and ends when the yolk is absorbed, fish are at different stages
of development with regard to such features as pigmentation,
eye development, and fin formation during this stage. The strik-
ing pigment rearrangements that occur during this stage provide
further emphasis that the yolk-sac stage is a transitional stage
between the egg and larval stages.
The Larval Stage
During the larval stage many ontogenetic changes occur (Mos-
er. 1981). Some of these relate directly to the development of
the adult form while other changes and structures are specialized
KENDALL ET AL.: ELH STAGES AND CHARACTERS
19
B
Fig. 10. Apparent convergence in siphonophore-mimicking appendages on larval fish. (A) Loweina rara. 17.6 mm. Note lower pectoral fin
ray (Moser and Ahlstrom, 1970); (B) Carapussp., 3.8 mm (Padoa, 1956j). Note elongate dorsal fin ray; (C) Exterilium larva, 64 mm. Note trailing
gut (Moser, 1981); (D) Lopholus sp., 12.t mm. Note elongate dorsal and pelvic ray (Sanzo. 1940); and (E) Arnoglossus japonkus, 30.5 mm.
Note elongate dorsal ray (Amaoka, 1973).
and of presumed functional significance primarily for planktonic
existence (Fig. 9). These latter features are of particular interest
in systematic studies of larval fish ontogeny. They include pig-
ment pattern, larval body shape, armature on head bones, and
precocious (early forming), elongate, or serrate fin spines. The
sequence and way of developing adult structures, such as the
skeleton and fin rays, are also useful larval characters. All of the
characters of the larvae— whether they are specialized larval
characters or merely characters observable in the larvae— may
have potential systematic value at some taxonomic level; how-
ever, the usefulness of most of the characters has not been eval-
uated (Tables 3 and 4).
Among the most taxonomically useful larval characters, gen-
erally at the specific or generic level, is the pigment pattern.
Usually, each species has a distinct larval pigment pattern. In
some the number and placement of individual melanophores
are diagnostic, while in others the location, shape, and size of
groups of melanophores are key characters. At a higher taxo-
nomic level, in the myctophiforms for example, the peritoneal
pigment blotches seem to indicate relationships on a suborder-
family level. Problems associated with the usefulness of pigment
patterns include 1 ) the widespread distribution of some patterns,
and 2) the variable state of melanophore contraction on larvae
of the same species. An example of the first problem is the
frequent occurrence of a row of small melanophores along the
ventral midline from just behind the anus to the tip of the tail.
Another example is a pigmented area midlaterally on the caudal
peduncle which occurs in numerous groups. A ventral spot at
the junction of the cleithra is also quite common. These are just
a few examples of widespread, presumably convergent pigment
patterns that limit the usefulness of pigment in systematic stud-
ies of larvae. The causes for the observed differences in degree
20
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 11. Liopropoma sp., 11.0 mm. Collected by G. R. Harbison,
16 May 1981, 6°31.8'S, 150°21.8'E. Note elongate dorsal spines.
of contraction of melanophores are not well understood al-
though they may be partially related to ambient light intensity.
The relative size and placement of melanophores are genetically
determined and therefore useful in a systematic context, while
the degree of contraction seems to be physiologically deter-
mined.
In general, the body shape and size at various stages of de-
velopment are characteristic of larvae at the generic or familial
level, although subtle differences in body shape may be char-
acteristic of species. Size at stage of development can be envi-
ronmentally modified (e.g., by temperature or food) to some
extent, but is primarily genetically determined. There appears
to be some convergence in larval body shape, such as on a long
tubular body in several divergent groups (e.g., Clupeiformes,
Argentinidae, Blennioidea), just as there is on the "herring"
morph of adults.
A valuable and fairly widespread set of larval characters con-
cerns the development of spines and armature on bones of the
head and cleithral region. Such armature has provided diag-
nostic larval characters as well as material for systematic infer-
ence at levels from species to order. Larval head armature ap-
pears to be a mark of the Acanthopterygii. Only a few scat-
tered examples of such armature appear in fishes which have
only soft rays as adults (e.g., Sudis). Within the spiny-rayed
fishes, beryciforms are quite heavily armed with spines on many
head bones. Perciforms usually do not have spines on the pa-
rietals but the supraoccipital is armed in some. The Scorpaeni-
formes are just the opposite: they tend to have head armature
that includes spines on the parietals but do not have spines on
the supraoccipital.
Nowhere are larval specializations more evident or varied
than in the fins. Elongation of particular spines or soft rays or
enlargement of whole fins are frequently seen. Such elongations
have been described for rays of the dorsal, pelvic, pectoral, and
caudal fins; thus they occur with both spines and soft rays. In
some, these long rays may bear pigmented "bulbs" or appear
like flagellae. Such specialized rays are produced in the dorsal,
pectoral, or pelvic fins of taxonomically diverse fishes. The ex-
tended gut of "exlerilium" ophidioid larvae (Fraser and Smith,
1974) and the serial pigment pattern of some leptocephali (Smith,
1979) may give the same appearance to potential predators as
these elongate rays. All of these structures may be mimicking
siphonophores: a remarkable example of convergence (Fig. 10
and 1 1 ). Elongate fin spines are heavy and armed with serrations
in some. Elongated rays are often precocious in development,
with some even forming in the egg. These fin characters seem
to vary at the family-species levels. Other characters associated
with fin development include the sequence of formation and
movement and loss of whole fins or some of the rays. Dorsal
and anal fins move forward along the body during larval de-
velopment in elopiform and clupeiform fishes. They develop in
"streamers" in the finfold of argentinoids and attach to the body
proper just before or during transformation. The shape of the
finfold, presence or absence of a preanal finfold, and shape of
the pectoral fin base provide additional characters at the family-
genus level.
Gut characters offish larvae include length and shape as well
as the development of a protruding, trailing hindgut in some.
In fishes with pholophores, their placement and sequence of
development are excellent characters at the subfamily-species
levels. The eye of a larva is specialized in a number of ways.
Fig. 12. Examples of special juvenile stages. (A) Hexagrammos lagocephalus. 28.0 mm. A neustonic or epipelagic form of a species that is
demersal as an adult (from Kendall and Vinter, 1984); (B) Forapiger longirosths. 17 mm. A spiny form that lives on tropical reefs as an adult
(from Kendall and Goldsborough, 1 9 1 1 ); (C) Sehaslolobus altivetis, 26.8 mm. A barred pelagic form of a species that is demersal on the continental
slope as an adult (from Moser et al., 1977); (D) Oncorhynchus kisulch. 37 mm. The freshwater alevin or parr stage of an andromous salmonid
(from Auer, 1982); and (E) Kali macrodon. 45 mm. The juvenile of a bathypelagic species. Originally described as Gargaropteron pterodactylops
(see Johnson and Cohen, 1974).
KENDALL ET AL.: ELH STAGES AND CHARACTERS
21
22
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Its size and rate of development are useful, as well as whether
it is round or oval. Some fish larvae have eyes borne on stalks
that reach an extreme in Idiacanthus, while others develop an
area of choroid tissue. Migration of the eye in flatfish larvae
from a symmetrical position to one side of the head is well
known. The sequence of development of ossified structures is
proving to be a powerful tool in systematic studies offish larvae.
The losses and fusions of bones, which are generally assumed
based only on adult material, can and should be tested using
developmental studies. The caudal fin skeleton has provided
excellent developmental characters to be used for systematic
inferences, mainly at the order-generic levels. The development
of scales has been little studied but may prove valuable, espe-
cially in fishes with precocious scales (e.g., some anthiins, hol-
ocentrids).
The Transformation Stage
Between the larval and juvenile stages, there is a transitional
stage which may be abrupt or prolonged and which, in many
fish, is accompanied by a change from planktonic habits to
demersal or schooling pelagic habits (Fig. 12). In some fishes
migration to a "nursery" ground occurs during or just before
this stage. Morphologically the transformation stage is charac-
terized by a change from larval body form and characters to
juvenile-adult body form and characters. At the end of this stage
the fish generally looks similar to the adult, with major differ-
ences only in pigmentation patterns. Two ontogenetic processes
occur during this stage of transition between the larva and ju-
venile: I ) loss of specialized larval characters, and 2) attainment
of juvenile-adult characters. Changes that occur during this stage
include pigment pattern, body shape, fin migration (e.g., in clu-
peids and engraulids), photophore formation, loss of elongate
fin rays and head spines (e.g., in epinepheline serranids and
holocentrids), eye migration (pleuronectiforms), and scale for-
mation.
In several groups, where the transformation stage is pro-
longed, the fish have developed specializations that are distinct
from both the larvae and juveniles. This stage has been desig-
nated the prejuvenile stage (Hubbs, 1943). The specializations
generally involve body shape and pigmentation. In many, the
morph resembles a herring-like fish and is apparently adapted
for neustonic life. The dorsal aspect of the fish is dark green or
blue and the lateral and ventral is silvery or white. The body
tends to be herring shaped and the mouth terminal. Fins are
generally unpigmented. Such a stage is present m Gadiformes
(Urophycis), Beryciformes (Holocentrus), Perciformes (e.g., Po-
malomus, MuUidae, Mugilidae) and Scorpaeni formes (e.g.,
Scorpaenichthys, Hexagrammos). In other fishes, such as some
myctophiforms and carapids, the prolonged transformation stage
may have distinctive body and fin shapes.
Implications of Larval Fish
Morphology
When studying the appearance of larval fishes, one is im-
mediately struck with their diversity and morphological dissim-
ilarity to adults. This dissimilarity led early workers to establish
names for several of these forms, not realizing that they were
the young stages of known adults. After establishing the identity
of many fish larvae in a variety of groups, we hypothesize that
the larvae of all species are recognizably distinct. The use of
diversity of larval form in vertebrate systematics was discussed
some time ago by Orton (1953b, 1955c, 1957) and in this vol-
ume we examine this use in detail in numerous groups of fishes.
Why are the larvae so diverse?— Despite the tremendous mor-
tality associated with living in the planktonic realm during the
larval period, survival must be sufficient to maintain the species
and provide a dispersal mechanism for it. To different degrees,
various taxa apparently rely on survival and longevity of in-
dividual larvae. The amount of reliance is presumably related
to fecundity and importance of dispersal and colonization to
the taxon. A number of structures have evolved that would be
expected to enhance larval survival in the plankton. Practically
no experimental work has been done to investigate the function
of larval structures, but some structures probably assist flotation
and feeding while others decrease predator mortality. Conver-
gence on characters that are apparently functionally important
to larval survival in the plankton is seen. These specializations
develop in conjunction with the basic ontogeny of the taxon.
In studying systematics using larval fishes, both the basic pattern
of development and the specialized structures must be analyzed.
Why are these larvae so morphologically unlike the adults?—
Most larvae are adapted to survive in an ecological realm (gen-
erally the plankton) that is far different from that of the adult.
These are small organisms, compared to adults, and they live
in the plankton, having to find and capture food there and avoid
becoming food. They float and migrate vertically in a milieu
that may be moving much faster than they are. During this
larval period, these fish undergo extreme changes in morphology
yet remain a functioning (eating, avoiding predators) organism
and eventually end up in a suitable nursery area for the juvenile
stage.
How then can larval morphology help us understand the evolu-
tion of these fishes?— Mler recognizing that each species has a
morphologically distinctive larva, generally we see that species
of the same genus are phenetically similar, and larvae of mem-
bers of a family also share common features. Even larvae of
suborders and orders share some larval characters. This would
be expected since evolution operates on all stages in the life
cycle, not just the adult. Evolutionary pressures on the larval
stage seem to be particularly intense in those groups that rely
on the larvae for widespread dispersal in the ocean. Here the
larvae appear well adapted for life in the planktonic realm, and
it can truly be said that the larva and the adult perform in "two
quite separate evolutionary theaters" (Moser and Ahlstrom,
1974). In this volume we are focusing on what we know to date
about larval evolution within various groups of fishes (Table 4).
Northwest and Alaska Fisheries Center, 2725 Montlake
Blvd. E., Seattle, Washington 98112 and Southwest
Fisheries Center, P.O. Box 271, La Jolla, California
92038.
TECHNIQUES AND APPROACHES
Early Life History Descriptions
E. M. Sandknop, B. Y. Sumida and H. G. Moser
FISHERIES studies require accurate identification of subject
species. Identification of the developmental stages of fishes
is complicated by the small size of the specimens, their fragility,
and the relatively great changes in their structure and pigmen-
tation. Experience has shown that major changes can occur over
very small growth increments and these can only be documented
by a continuous growth series. Published descriptions of de-
velopmental series vary in quality, perhaps more than do species
descriptions of adults. Prior to Bertelsen (1951) and Ahlstrom
and Ball (1954), most published descriptions were based on
relatively few specimens, which were described individually. In
their study of the early life history stages of the jack mackerel
(Trachunts syinmetricus), Ahlstrom and Ball (1954) used over
500 eggs and a series of about 250 larvae, transforming speci-
mens, and juveniles to describe development. Changes in struc-
ture and pigmentation were thus described as a dynamic con-
tinuum, with emphasis on variation, in contrast to the approach
of most previous workers. Developmental osteology was con-
sidered an integral part of the description as were seasonal and
geographic distributions of eggs and larvae. This paper was fol-
lowed by several others (Ahlstrom and Counts, 1955, 1958;
Uchida et al., 1958; Kramer, 1960) and these became models
for subsequent descriptive papers, including some which treated
several species in various taxonomic groups (Moser and Ahl-
strom, 1970; Ahlstrom, 1974; Ahlstrom et al., 1976; Moser et
al., 1977; Kendall, 1979; Brownell, 1979; Richardson and
Washington, 1980; Fahay, 1983; Leis and Rennis, 1983). The
following is a brief account of the elements involved in preparing
early life history accounts of teleosts.
Sources
The major source of material is plankton collections. Typical
survey tows strain a column of water 200 m to the surface and
sample eggs and subsequent larval stages of a major portion of
the fish fauna (Smith and Richardson, 1 977). Fishes which have
highly stratified vertical distributions are undersampled by
oblique tows and require special gear or tow strategies. For
example, surface dwellers can be sampled by neuston nets (Zait-
sev, 1970; Nellen and Hempel, 1970; Hempel and Weikert,
1972; Nellen, 1973a; Ahlstrom and Stevens, 1976) and those
species residing near the bottom may be sampled by epi-benthic
plankton nets (Schlotterbeck and Connally, 1 982). Larger larvae
and transforming stages are poorly sampled by typical survey
tows principally because of accumulated mortality, increased
avoidance capacity, and migration out of the sampling zone.
These stages are more effectively sampled by trawls (Tranter,
1968), dip-netting with attractor lights (Klawe, 1 960), light traps
(Faber, 1982), and fish predators (Haedrich and Nielsen, 1966).
Recently, scuba divers have collected oceanic larvae with their
delicate structures intact (Harbison et al., 1978; Govoni et al.,
1 984). Developmental series may also be obtained by rearing
larvae from eggs collected at sea or from captive brood stock
(Houdeetal., 1970, 1974; Houde and Swanson, 1975; Richards
etal., 1974; Houde and Potthoff, 1976; Moser and Butler, 1981).
This method becomes essential when working with speciose
faunas (e.g., Sebastes, warm water shorefishes), if only to de-
termine which species cannot be identified.
Use of Specimens
The characters and techniques used in identifying develop-
mental stages are discussed elsewhere in this volume (see Ken-
dall et al.; Matarese and Sandknop; Powles and MarkJe). From
the continuous developmental series two subseries are assem-
bled and these form the basis for the description. The first series
is used to describe morphology and pigmentation. Specimens
in the second series are cleared and stained by a variety of
techniques to describe the development of cartilaginous and
osseus features (Potthoff, this volume).
The number of specimens used to construct these series is
dependent on several factors: 1) specimen availability, 2) length
(duration) of the development period, and 3) complexity of
developmental change. A guideline is that there should be enough
specimens to demonstrate the beginning, progression and com-
pletion of significant developmental changes in morphology and
pigmentation. Usually more specimens are required for species
which have extended larval periods; however, many fishes which
transform at small sizes undergo great change over small length
intervals. For example, lined sole {Achirus lineatus) hatch at 1 .6
mm, transform at about 4.0 mm, and complete a large suite of
developmental changes over a 2.5 mm length interval (Houde
et al., 1 970). The majority of marine teleosts transform between
10 and 30 mm and, for these, major developmental events can
be documented by specimen length increments of 0.5-1.0 mm.
Multiple samples representing 1 mm-intervals are required to
study fine-scale character variation; however, such studies have
rarely been done (Ahlstrom and Moser, 1981).
A table of morphometric measurements constructed from the
unstained series provides data on the size at important devel-
opmental milestones (e.g., hatching, notochord flexion, fin for-
mation, transformation) and provides a basis for analyzing
structural change and allometric growth. These specimens can
be used to construct character matrices of complex or diagnostic
pigment changes. Illustration specimens chosen from the series
provide an integrated view of major characters and also, if ac-
curately executed, are themselves morphometric and meristic
documents (Sumida et al., this volume).
The stained series is used to construct a meristic table that
forms the basis for following the development of fin rays and
supporting elements, the axial skeleton and cranial bones (Dunn,
this volume). Fine bony structures, such as cranial spines are
also apparent in these preparations.
Published descriptions employing these basic elements are
23
24
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
the basis for ontogenetic studies of fishes. These are essential
for the identification of ichthyoplankton collections, and also
present characters for systematic analysis. Data provided in
these descriptions have proved useful in studies of the physi-
ology, behavior and ecology of the early stages of fishes.
National Marine Fisheries Service, Southwest Fisheries
Center, P.O. Box 271, La Jolla, California 92038.
Synopsis of Culture Methods for Marine Fish Larvae
J. R. Hunter
THE objective of this paper is to provide a synopsis of present
technology for small-scale laboratory culture of marine fish
larvae. The technology of marine fish culture is relevant to this
book because it is one of the best ways to obtain a taxonomic
series. "Ahlie" Ahlstrom was a strong proponent of this ap-
proach and I lectured on the subject at his request for his courses
on larval fish systematics. Marine fish culture has often been
reviewed (May, 1970-, Houde, 1972a; Houde and Taniguchi,
1979; Shelboume, 1964; Kinne, 1977) and many additional
references may be found in the previous reviews. The key feature
of my review is that it is a condensed practical guide and key
to the literature for beginners interested in small-scale laboratory
culture of marine fish larvae; culture of freshwater fishes is not
considered.
Eggs
Sources. — Pelagic fish eggs can be obtained from plankton tows,
by catching ripe fish and fertilizing the eggs, and by induction
of spawning of laboratory brood stock.
Let eggs taken in plankton tows stand in quart bottles for 0.5
h, then remove plankton from bottom of jar and add fresh sea
water (a second decanting may be required). Jars are stored on
their sides in an insulated ice box with a refrigerant for 24 h or
longer with the temperature kept within spawning range.
Virtually all marine clupeoid fishes (Blaxter and Hunter, 1982)
and probably most other pelagic marine fishes spawn at night,
hence running ripe fish are more common at night or just before
sunset (final egg maturation or hydration occurs just before
spawning). After an egg is spawned in sea water its fertility
decreases but the maximum time for it to become infertile is
highly variable among species, varying from 6 minutes to over
3 hours (Ginzburg, 1972). Sperm in sea water may remain fertile
for days (Ginzburg, 1972) although fertility periods as short as
30 seconds have been observed (Haydock, 1971). Owing to the
great variation in the time eggs and sperm remain fertile it is
preferable that sperm and eggs be mixed immediately after they
are obtained.
Storage of gametes may be helpful since mature males and
females are not always available simultaneously and crosses
between subpopulations may be desired. It is well known that
sperm can be stored for extended periods ( 10 or more hours) if
kept cool and maintained in the concentrated form and not
activated by sea water (Ginzburg, 1972; Erdahl and Graham,
1980). Fertilization of Clupea harengus eggs may be obtained
after 6-7 days dry storage at 4° C but a high hatching rate is
expected only after periods less than 36 h (Blaxter and Holli-
day, 1963). It is now possible to extend the life of fish sperm
for much longer periods using cryopreservation techniques
(- 196°C) (Erdahl and Graham, 1980). Various cryoprotective
agents have been used to freeze sperm of marine fishes including
glycerol (Blaxter and Holliday, 1963), glucose, NaCI, Ringer's
solution and fish serum (Hara et al., 1982).
The stress of capture causes female Katsiiwonus pelamis to
ovulate and spawn within 24 h after capture but eggs are often
not viable (Kaya et al., 1982), Maturing marine fish in the lab-
oratory and spawning them by hormone injections has become
routine in recent years and is preferable to stress techniques.
Examples include Engraulis mordax (Leong, 1971), Scomber
japonicus (Leong, 1977), Chanos chanos (Liao et al., 1979),
Bairdiella icistia (Haydock, 1971), Paralichthys denial us and
Pseudopleuronectes americanus (Smigielski, 1975a, b) and oth-
ers (see review of Lam, 1982). Induction of spawning in the
laboratory may require an open sea water system, large holding
tanks (e.g., -3 m dia. or larger), temperature and light control.
Handling and stocking.— To count eggs without damaging them
we recommend a polished wide bore (~3 mm) pipette; count
30-50 late stage eggs at a time in a depression slide under a
dissection microscope, and wash eggs off the slide by immersion
of the entire slide in sea water. Counting eggs is critical because
higher mortalities and slower growth result from excess stocking
densities (Houde, 1975 and 1977). As a rule stocking densities
in rearing tanks of 8 eggs/I or less seems preferable and most
rearing successes have occurred when stocking did not exceed
20 eggs/1 (Houde, 1975). Similarly, the mortality of Mugil ceph-
a/(« larvae seems to remain constant (2-3% loss/day) at stocking
densities of 1-30 larvae/1 (Kraul, 1983).
Apparatus
Containers and lighting. — Larvae appear to grow faster and show
fewer signs of starvation when reared in large containers (100
1) rather than in smaller ones (10 1) (Theilacker, 1980b). Opti-
mum container size doubtless varies with species but 40 1 con-
tainers are probably the minimum size that should be used and
I prefer 100-400 1 containers. We use cylindrical black fiberglass
containers although excellent results are obtained using ordinary
rectangular glass aquaria (Houde, 1975).
It is traditional to provide a daily cycle of illumination to
HUNTER: CULTURE METHODS
25
larvae in rearing containers although constant illumination is
occasionally used. Typically fluorescent lamps are used which
provide 2,000-3,000 lux at the water surface (Houde, 1978;
Hunter, 1976). Night light levels vary; we provide no light at
night whereas Houde (1978) provides a dim light of 40-90 lux
at night, which is substantially above the visual threshold for
feeding for larval E. morda.x (6 mm larvae 50% feeding thresh-
old = 6 lux, and 10-15 mm larvae 50% threshold = 0.6 lux,
Bagarinao and Hunter, 1983). Clearly, longer periods for visual
feeding will probably enhance growth if food is limited. Rearing
at high light intensities such as natural sunlight may greatly
increase production of algae and zooplankton in the culture tank
and thereby increase larval survival (Kraul, 1983). On the other
hand, solar UV radiation is clearly lethal to younger larvae
(Hunter etal., 1 982) and use of deep tanks, or shaded or covered
tanks (screen cloth, acrylic plastic, glass or mylar film) is rec-
ommended for the first 1-2 weeks of larval life if tanks are to
be exposed to solar radiation.
Water qualily.—C\osed, non-circulating systems are typically
used to rear marine fish larvae at least during the younger stages,
because in an open system planktonic larvae and their foods
are easily lost. Older (nektonic) larvae are able to resist the
current and to consume a daily ration in a short period so a
partially open system can be used. We fill our rearing containers
with UV treated sea water that is passed through three, in line,
cartridge filters (5, 3 and 1 ^m pore).' Although not a common
practice in small scale rearing work, the addition to rearing tanks
of antibiotics (sodium penicillin G at 50 i.u./ml plus strepto-
mycin sulphate at 0.05 g/ml) slightly improved survival of Pleu-
ronectes platessa eggs through hatching, but surprisingly this
single treatment greatly improved survival of larvae through
metamorphosis (Shelboume, 1975).
Use of a closed system requires attention to water quality, a
problem which may be intensified at higher rearing tempera-
tures. In the most complete study of water quality in rearing
tanks for marine fish larvae, Brownell (1980a, b) considered
seven variables (pH, dissolved oxygen, carbon dioxide, am-
monia, nitrite and nitrate), but only high pH, low dissolved
oxygen and un-ionized ammonia had effects at levels likely to
be encountered in rearing tanks. First feeding incidence declined
by 50% in all species he studied when dissolved oxygen con-
centrations were between 4 and 4.75 mg/1 (49-58% saturation).
Dissolved oxygen in our rearing containers usually is not sat-
urated after planktonic foods are added, and typically it is about
80% saturation even with aeration. Clearly water quality is im-
proved by aeration and frequent water changes and lank clean-
ing. Werner and Blaxler (1980) exchanged 20% of the water in
Clupea harengus cultures (9° C) 3 times per week but at high
temperatures greater replacement rates are required. For ex-
ample Houde (1977) replaced 20% of the tank sea water on
alternate days while culturing Anchoa mitchilli and Achirus lin-
eatus at 26-28° C. Frequent tank cleaning is important as heavy
mortalities may result from toxins produced by debris on the
container bottom (Kraul, 1983). Aeration, unless very gentle,
can cause heavy mortalities among delicate eggs and newly
hatched larvae. In fact, Shelboume (1964) recommends no aer-
' Aqua-Pure model APIO. AMP Cuno Division, Inc., Meriden. Con-
necticut USA.
ation for Pleuronectes platessa larvae. I recommend very gentle
aeration but not until a week or so beyond the first feeding stage.
The mortality of cultured fish larvae often increases during
the period of initial swim bladder inflation in physoclistous
fishes (Doroshev et al., 1981; Kuhlmann et al., 1981) and this
could be related to water quality. Symptoms include delay or
complete failure of inflation or excessive inflation; in either case
normal swimming patterns are disrupted and death frequently
results. The causes of abnormal inflation are not clear; preven-
tion of larvae from reaching the water surface prevented excess
inflation in M. cephalus larvae (Nash et al., 1977), whereas the
same treatment in Atractoscion nobilis larvae had no effect. In
A. nobilis excess inflation was associated with abnormal devel-
opment of gas secretory tissue suggesting a more complex etiol-
ogy (SWFC. unpubl. data). Failure to inflate the swim bladder
is a common problem in Morone saxatilus culture and turbulent
aeration may reduce the incidence of this disease (Doroshev and
Comacchia, 1979) but it now appears that reduction in salinity
from 17 ppt to 4 ppt has a much greater eflect in reducing the
incidence of swim bladder malfunction (S. Doroshev and J.
Merritt, U. Cal. Davis, pers. comm.).
Food
The most critical aspect of rearing marine larvae is manage-
ment of their food. Food must be the correct density, size,
nutritionally adequate and must remain suspended in the water
column which usually requires the use of living pelagic organ-
isms.
Food size.— Typ\c&\ pelagic fish larvae are 2.5-4.0 mm when
they begin feeding and acceptable prey are 20-1 50 /um in breadth
(Houde and Taniguchi, 1979). Some large larvae, e.g.. larval C.
harengiis (B\di\\.QT. 1965). Pleuronectes platessa {Riley. 1966) or
small larvae with large mouths, e.g., Merluccius productus {Sum-
ida and Moser, 1980), can begin feeding on prey 300 Mm or
larger in breadth. The optimal food size increases as larvae grow
(Hunter, 1981), so any culture technique should provide a stead-
ily increasing range of food sizes, because if the food is too small
growth slows and mortality occurs (Hunter, 1981). Food size
requirements can be expressed in terms of the ratio of prey width
to mouth width. The 50% threshold for feeding on a prey of a
particular width occurs when this ratio is about 0.75, although
occasionally larvae consume prey as wide as the width of their
mouth (ratio = 1) (Hunter, 1981). At the onset of first feeding
a small prey of about 'A the mouth width seems to be preferable
as capture success is low at this time but within a few days larvae
are able to consume food of about V2 the mouth width.
Wild zooplankton— V/i\d zooplankton, primarily the naupliar
and copepodite stages of marine copepods but also mollusc
veligers, tintinnids, cladocera, and appendicularia larvae, are
the natural foods of most marine fish larvae and probably also
the best source of food for rearing a larval taxonomic series.
Wild zooplankton provide a wide range of sizes and types and
are probably nutritionally superior to cultured rotifers and Ar-
lemia nauplii (Kuhlmann et al., 1981). Collection of wild zoo-
plankton may require less effort than production of cultured
food except for brine shrimp nauplii (see below). Zooplankton
is collected in nets of about 50 ^m, and is graded by size in the
laboratory using various nylon nets (Houde, 1977, 1978), This
eliminates the larger zooplankton which larvae would be unable
26
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
to consume and which may be larval predators. Fish larvae,
particularly yolk-sac stages, are vulnerable to various carnivo-
rous copepods, amphipods, euphausiids and chaetognaths
(Hunter, 1981).
Cultured foods.— T-wo cultured foods, the rotifer Brachiomts
plicatilis, and nauplii of the brine shrimp, Arteinia. should be
considered as potential foods for rearing marine fish larvae as
many fish larvae can be reared on a combination of these two
foods. These two foods may also be used as a supplement to
diets of wild plankton. Groups of fishes that have been reared
to metamorphosis on a combination oi Brachionus and Anemia
or on Artemia alone include C. harengns, species of serranids,
scombrids, atherinids, various flatfishes, sciaenids, and saganids
(May, 1970; May etal., 1974; and unpubl. SWFC data). /lr?ew;a
nauplii are recommended only for larvae with differentiated guts
as they are quite resistant to digestion whereas copepods are not
(Rosenthal, 1969).
Methods for culturing rotifers using algae are given by Thei-
lacker and McMaster (1971); culture methods employing for-
mulated artificial diets or freeze dried algae (Gatesoupe and
Robin, 1981; Gatesoupe and Luquet, 1981) and ones using
brewers yeast also exist. Many of the essential facts given in
these original papers will not be repeated here but I will point
out a few practical points regarding rotifer culture using algae.
Suitable algae species for rotifer culture include Dunaliella,
Nannochloris, Tetraselmis, and Chlorella which may be grown
using standard culture media (Guillard, 1975) or using liquid
commercial plant fertilizers (dosage for fertilizer containing 8%
total nitrogen = 0. 1 ml of fertilizer/1; dosage among brands is
adjusted depending on total N content). We prefer commercial
plant fertilizers that have an organic base such as liquid fish
fertilizers and avoid those that have soil penetrants. A daily
doubling rate can be expected in healthy rotifer cultures, and
cultures can be maintained for weeks or even months by adding
fresh algae or nutrients and sea water, although single batch
harvesting after about 2 weeks gives more dependable results.
Rotifers are harvested using gravity flow through a nylon filter
(20-40 ^m mesh) as pumps may kill rotifers.
Production ofArlemia nauplii is simple since all that is needed
is to hatch the cysts ("Anemia eggs"). Cysts from a variety of
strains of Anemia are commercially available. The strains differ
considerably in average naupliar size (423-775 ^m length), in
pesticide content (DDT, PCB, and chlordane) and in certain
fatty acids (Klein-MacPhee et al., 1982). These authors show
that very low survival (15%) of P. amehcanus larvae occurred
when they were fed San Pablo Bay (San Francisco) nauplii
whereas survival of larvae fed other strains varied from 60-
80%. Beck et al. ( 1 980) gave similar results for Menidia menidia
larvae. Of all the strains tested in these papers the Australian
and Brazilian strains seem the most suitable for rearing larvae
and the San Pablo Bay (USA) the least. -
Anemia hatcheries vary from a jar to complex automated
systems. The J. D. Riley Anemia hatching box has been used
with slight modification in many laboratories for over 20 years.
It is a sea water filled box separated in half by a sliding partition;
Anemia cysts are added to one side (I g/l) and they hatch 1-2
^ Exotic Anemia cysts are available from: Artemia Inc., P.O. Box
2891, Castro Valley, California 94546 USA and Biomarine Research.
4643 W. Rosecrans, Hawthorne, California 90250 USA.
days later depending on the temperature selected (23-30° C).
The tank is then illuminated, the partition raised slightly off the
bottom, and the nauplii, attracted by the light, swim beneath
the partition leaving behind the hatching debris and unhatched
cysts (Shelboume, 1964). A semiautomatic version of this sys-
tem is described by Nash (1973), and various other improve-
ments in aeration, illumination, temperature, and other factors
have increased yields to lO' nauplii per 4.8 g of cysts (San
Francisco Bay Brand) (Dye, 1 980). In recent years decapsulation
of Anemia cysts using hypochlorite bleach has become popular
because it increases yields, increases the dry weight of the nau-
plius (Bruggeman et al., 1 980) and eliminates contamination of
larval fish rearing tanks with unhatched cysts.
It should also be noted that freshly hatched Anemia nauplii
are clearly more nutritious than older starving individuals and
consequently new batches should be frequently produced. In
general, prey with full stomachs are probably nutritionally pref-
erable to ones with empty stomachs. Similarly, more Dicen-
trarchits labrax larvae seem to survive when rotifers are nutri-
tionally enhanced by 30 min immersion in a solution containing
vitamins and soluble proteins (Gatesoupe and Luquet, 1981).
Mass culture of marine copepods is difficult and laborious
and therefore not recommended when a taxonomic series is the
sole objective. Nevertheless, culture of marine copepods may
be the only way some fish larvae can be reared if wild zooplank-
ton is not readily available and larvae die when fed Anemia
nauplii (rarely are more than a single strain of Anemia tested,
however). Harpacticoid copepods (Tignopus sp., Tishe sp., and
Euterpina sp.) are the most frequently used copepods because
of ease of culture; for culture techniques see Kahan et al. (1982)
and Hunter (1976). Euterpina may be preferable to Tignopus
or Tishe because the nauplii and copepodites of Euterpina are
pelagic and therefore available to the larvae whereas nauplii and
copepodites of Tigriopus and Tishe tend to remain on surfaces
and are therefore less available (Kraul, 1983). See Nassogne
(1970) and Zurlini et al. (1978) for laboratory culture of Euter-
pina.
Eood density. —The optimal food density for fish larvae depends
upon the size of the food organism and size or age of the larvae.
Densities of 1-3 organisms/ml have been routinely used for
larvae fed wild zooplankton (largely copepod nauplii) during
the first 1-2 weeks of feeding (Houde and Taniguchi, 1979).
The same density range is used when cultured .Anemia nauplii
are the food. A higher density range (IO-20/ml) is used for
cultured B. plicatilis which are about 1/10 of the weight of an
.irtemia nauplius (Theilacker and McMaster, 1971). A very
small food particle, the dinoflagellate Gymnodinium splendens
(40 nm dia), is used for the first 2 days of feeding in northern
anchovy larvae (Lasker et al., 1970; Hunter, 1976) at a high
density of about lOO/ml. In very active species such as S. ja-
ponicus or the siganid Siganus canaliculatus high food densities
can cause heavy mortality because of overfeeding since most
larval fishes seem to lack a satiation mechanism (May et al.,
1974; Hunter, 1981). Overfeeding seems to occur only when
such easily captured prey as .irtemia nauplii are used as food.
Piscivorous fish /arvac — Piscivorous fish larvae such as the
scombroids, Sphyraena and others pose special problems in
culture. Fish larvae are an ideal food for such larvae; in fact,
our only success in rearing Katsuwonus pelamis larvae to meta-
morphosis was probably related to an abundant supply of yolk-
HUNTER: CULTURE METHODS
27
sac fish larvae as food. Zooplankton is the initial food until
piscivorous feeding habits develop (Houde, 1972b; Mayo, 1973;
Hunter and Kimbrell, 1980). Piscivorous larvae manipulate their
larval prey and consequently are less dependent on mouth size
when consuming larval fish. Sibling cannibalism is common
under reanng conditions in such fishes. Increasing the food den-
sity may increase survival as may elevating the temperature,
thereby accelerating growth through the most cannibalistic sizes;
at least in scombroids sibling cannibalism declines at meta-
morphosis (Mayo. 1973; Hunter and Kimbrell, 1980). Sorting
by size and isolating the larger larvae is probably the only certain
method for controlling losses due to cannibalism, however.
Phytoplankton
Phytoplankton blooms are often maintained in larval culture
tanks to reduce the detrimental effects of metabolic by-products
which accumulate in static rearing tanks (Houde, 1974) and to
provide food for larval food organisms. In many cases dense
blooms of phytoplankton enhance larval growth and survival
and I recommend the practice but the mechanism is obscure.
The phytoplankters used are various, easily grown, small species
such as Chlorella. Anacystis, Nannochloris, Tetraselmis. Dun-
aliella. Isochrysis. Phaeodactylum and others.' They are main-
tained at high densities (10,000 or more cells/ml) in the rearing
tanks. At high cell densities larvae ingest these small phyto-
plankters, perhaps inadvertently (Moffatt, 1981) but they appear
not to be able to exist on them as a sole food source (Houde,
1974; Scura and Jerde, 1977). They may supplement the food
' For a nominal fee starter cultures of manne phytoplankton can be
obtained from R. R. L. Guiliard. Bigelow Laboratory for Ocean Sciences.
McKown Point, West Boothbay Harbor, Maine 04575 USA; culture
methods are discussed by Guiliard (1975).
ration either directly or indirectly through the ingestion of prey
having guts full of algal cells (Moffatt, 1981). Evidence now
exists that enhancement of growth and survival of larval Scoph-
ihalmus maximiis by blooms of Isochrysis and Phaeodactylum
is due to the inclusion in the diet of certain polyunsaturated
fatty acids not occurring in the normal laboratory rotifer diet
(Scott and Middleton, 1979). It is interesting in this regard that
Dunaliella which lacks the fatty acids did not enhance S. max-
imiis larval growth or survival.
Effects of Culture
Extrapolation from cultured larvae to natural populations must
be done with caution because culture may affect the morphology,
behavior and biochemistry of larvae (Blaxter, 1976). The mor-
phological characteristics most susceptible to modification in
tanks are those partially controlled by environmental conditions
such as vertebrae and fin ray counts. Reared larvae also may
be more heavily pigmented than sea caught specimens (Watson,
1982). This appears to be related to the expanded nature of the
melanophores, not to added numbers of pigment cells. In ad-
dition, pigmentation events may occur at smaller sizes in reared
material (S. Richardson, Gulf Coast Research Laboratory, Ocean
Springs, Mississippi, pers. comm.). Laboratory reared larvae are
often heavier and have deeper bodies than their wild counter-
parts, making some morphometric measurements on laboratory
specimens useless (Blaxter, 1975). The differences in preserva-
tion and handling between laboratory and sea-caught larvae also
make direct size-specific comparisons difficult. Shrinkage in
length may vary greatly depending on the duration larvae re-
main in plankton nets and shrinkage differences between reared
and wild specimens can be misinterpreted as morphological
differences (Theilacker, 1980a).
National Marine Fisheries Service, Southwest Fisheries
Center, P.O. Box 271, La Jolla, California 92038.
Identification of Fish Eggs
A. C. Matarese and E. M. Sandknop
A wide variety of egg types exists among teleost fishes in both
freshwater and marine environments. Eggs may be pelagic
and nonadhesive or demersal and either adhesive or not. They
may possess a variety of specialized structures aiding in flotation
or attachment. Depending on egg type and associated repro-
ductive ecology, many characters are useful in identification.
These characters have been reviewed for pelagic marine eggs by
Rass(1973), Robertson (1975a), Russell (1976), and Ahlstrom
and Moser ( 1 980); we have liberally and extensively drawn from
the latter. Important characters for other egg types have been
discussed in part by Balon (1975a, 1981a), Hardy (1978a, b),
Jones et al. (1978), and Snyder (1981). Characters such as size
and possession of oil globules are important for all types; how-
ever, perivitelline space and chorion sculpturing are more im-
portant in pelagic eggs, while in demersal eggs special coatings.
chorion thickness, or nature of egg deposition may be more
useful.
A wealth of potential characters useful in egg identification
exists; however, it is still difficult to identify eggs of most species
with certainty. Except for late stages, few may be recognized at
the species level. Some characters are useful at a family level,
but presently it is not productive to speculate on the systematic
significance of any characters (see Kendall et al., this volume).
Presently, the main goal of taxonomy with respect to fish eggs
is identification.
Regardless of egg type or reproductive ecology, a summary
of identification characters useful to an egg taxonomist is pre-
sented. Additionally, we recommend using available literature
for reference and encourage the building of local fish egg col-
lections. We follow Ahlstrom and Ball (1954) in subdividing
28
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
1.34x0.66
Engraulis mordax
B
1.0x1.06
Ophidion scrippsae
Unidentified
0.58-0.74
Vinciguerria lucetia
1.9
Glyptocephalus zachirus
0.80
Symphurus atricauda
H
Prionotus stephanophrys
2.92
Icosteus aenigmaticus
1.35
Etrumeus teres
Fig. 13. Fish eggs. Captions under each illustration indicate the species and the diameter or dimensions of the egg in millimeters. A. Engrauli
mordax. original; B. Ophidwn scrippsae. onginal; C. Unidentified, original; D. Vincigiierna tucetia. from Ahlstrom and Counts (1938); E
Glyptocephalus zachirus. from Ahlstrom and Moser (1980); F. Symphurus atricauda. original; G. Prionotus stephanophrys. onginal; H. Icostei.
aenigmaticus, original; and I. Etrumeus teres, original.
'is
E.
•osleus
MATARESE AND SANDKNOP: EGG IDENTIFICATION
29
egg development as follows: Early— from fertilization to closure
of blastopore. Middle— from closure of blastopore to tail bud
lifting off yolk, and Late — from tail bud lifting off yolk to time
of hatching.
Identification Characters
Shape.— The vast majority of all egg types are spherical. Ex-
ceptions include ellipsoidal eggs as found in anchovies, En-
graulis and Anchoa. and slightly flattened or ovoid eggs as seen
in members of the families Gobiidae, Scaridae, and Ophidiidae
(Fig. 13A. B). A number of demersal eggs have somewhat ir-
regular shapes, especially those associated with large egg masses.
The perciform family Congrogadidae has cruciform shaped eggs
(Herwig and Dewey, 1982). An unidentified, star-shaped egg is
encountered infrequently in the Alaska region (Fig. 13C).
Size.—T\\t average marine and freshwater fish egg size is about
1.0 mm. According to Ahlstrom and Moser (1980), pelagic fish
eggs range from 0.5 mm [Mncigiicnia (Fig. 13D)] to about 5.5
mm (Muraenidae). Demersal eggs may range higher in size (up
to 7.0-8.0 mm), e.g., members of the families Salmonidae, An-
arhichadidae, and Zoarcidae. Mouth brooders, e.g., in the catfish
family Ariidae, have among the largest eggs with sizes from 1 4
mm to 26 mm.
Oil globules.— The oil globule provides useful characters in fish
egg identification; these include presence or absence, number,
size, position, color, and pigmentation. Among both pelagic and
demersal eggs, the most common form contains a single oil
globule. Eggs may lack an oil globule as in most gadines and
pleuronectids (Glyplocephaliis). contain only one (Icosteiis), or
have multiple oil globules as in the cynoglossids and triglids
(Symphums and Prionotus) (Fig. 13E, F, G, and H). In pelagic
eggs with a single oil globule, the size ranges from <0.10 mm
to > 1.0 mm (Ahlstrom and Moser, 1980). The position of the
oil globule within the yolk sac is usually posterior, but several
groups contain species that have an anterior placement (e.g.,
labrids and carangids) and others have an intermediate place-
ment (argentinids). In some fishes, oil globules migrate during
embryonic development. Some members of the family Bathy-
lagidae initially possess multiple oil globules that eventually
coalesce into a single globule (Ahlstrom, 1969). Although not a
totally reliable character, the oil globule color can be useful,
especially in the identification of freshly taken demersal eggs.
Lastly, many species have oil globules with melanistic pigment,
Icosteus (Fig. 13H) and Icichthys.
Yolk.— The degree of yolk segmentation is an important iden-
tification character. Yolk is usually segmented in primitive forms,
e.g., Etruineus (Fig. 131), and homogeneous in higher forms
(Rass, 1973; Ahlstrom and Moser, 1980). The opaqueness of
yolk found in catfishes, salmonids, and gars can be diagnostic'
Pigment, which may also be diagnostic, can be present dunng
various developmental stages from middle to late. Yolk color
is often important especially in demersal eggs. Among demersal
eggs vitelline circulation patterns within the yolk sac are useful
in identification.'
' P. Douglas Martin, Chesapeake Biological Laboratory, P.O. Box 38,
Solomons, Maryland 20688. Personal communication, October 1982.
Chorion. — A. number of characteristics associated with the cho-
rion or egg envelope can be useful in identifying fish eggs and
have been shown to be highly adapted to the environmental
conditions under which an embryo develops (Ivankov and Kur-
dyayeva, 1973; Stehr and Hawkes, 1979; Laale, 1980; Stehr,
1982). The most important character of the chorion is whether
it is smooth, as is in most fishes, or sculptured. Among fish eggs
with patterns, the size and texture (e.g., raised hexagons, pus-
tules) of the design are diagnostic. Raised polygonal surfaces are
found in several unrelated species (Stehr, 1982), e.g., Synodus
and Pleuronichthys (Sumida et al., 1979), and pustules occur
among some bathylagids and argentinids. Mugil cephalus eggs
(Fig. 14A), previously considered to have a smooth chorion,
have a raised patterned surface visible by scanning electron
microscope (Boehlert, this volume). In many groups of fishes,
the chorion has various degrees of ornamentation consisting of
projections, threads, filaments, or stalks which may aid in flo-
tation (pelagic) or attachment (demersal). In some scombere-
socids, e.g., Cololahis (Fig. 14B). some exocoetids and ather-
inids, pelagic eggs are attached to each other or to a substrate
by filaments. Spines are found in some myctophiforms and
exocoetids, and stalks occur in some demersal egg groups, e.g.,
blenniids and Osmerus mordax. In ostraciid eggs, a patch of
pustules is present near the micropyle (Fig. 14C).
Recently, thickness of the chorion has been of diagnostic value
(Ivankov and Kurdyayeva, 1973; Boehlert, this volume). Stehr
and Hawkes (1979), using scanning electron microscopy, found
that most marine teleosts with pelagic eggs have thin chorions
in relation to egg diameter whereas demersal eggs tend to de-
velop much thicker chorions. Color of the chorion is an im-
portant diagnostic character, especially for freshly taken de-
mersal eggs in the marine intertidal environment (Matarese and
Marliave, 1982). A number of freshwater demersal fishes have
eggs that possess a special coating associated with the chorion
which can be either gelatinous or adhesive, e.g., Perca. Icialurus,
and Notropis (Snyder, 1981).
Penvilelline space. — Most fish eggs have a narrow- to medium-
width perivitelline space, but wide spaces are common in some
groups, especially among the more primitive fishes that have a
segmented yolk, e.g., Clupeiformes (Sardinops. Fig. 14D), An-
guilliformes, and Salmoniformes (Chauliodus. Fig. 14E) (Ahl-
strom and Moser, 1980). Large perivitelline spaces are also found
among some unrelated higher forms, such as cypnnids (Nolro-
pi.s). percichthyids (Morone saxatill.s). or pleuronectids (Hip-
poglossoides).
Embryonic characters.— CharacXers associated with the devel-
oping embryo are extremely useful in egg identification, partic-
ularly in the middle and late stages of development. Many eggs
not identifiable in the early stages are easily recognizable using
embryonic characters such as pigment on embryo or finfold and
morphology. In some fishes, embryonic pigment in the late stages
has already undergone sufficient migration and rearrangement
to the point where it resembles the yolk-sac larva; this is com-
mon in several groups including gadiformes, e.g., Merluccius
(Fig. 14F), Gadus. and Theragra. and heavily pigmented flat-
fishes like Pleuronichthys and Hypsopsetta. Characteristic late-
stage pigment bands appear in Glyptocephalus (Fig. 13E). In
most freshwater species, pigment is not present prior to pigment
cell migration but appears sometime after the cells have mi-
30
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
0.76-0.80
Mugil cephalus
B
1.7x1.9
Cololabis saira
1.54x1.68
Ostraciidae
1.35-2.05
Sardinops sagax
2.93
Chauliodus macouni
1.07-1.18
Merluccius productus
H
2.0
Eumicrotremus orbis
2.65-2.90
Trachipterus altivelus
0.88
Stomias atri venter
Fig. 14. Fish eggs. Captions under each illustration indicate the species and the diameter or dimensions of the egg in millimeters. A. Mugil
cephalus. original; B. Cololabis saira. original; C. Ostraciidae, original; D. Sardinops sagax. original; E. Chauliodus macouni. original; F. Merluccius
productus. from Ahlstrom and Counts ( 1 955); G. Eumicrotremus orbis. from Matarese and Borton unpubl. MS; H. Trachipterus altivelus. original;
and I. Stomias aim-enter, original.
grated lo their actual destinations (Snyder, 1981). As seen in
the cyclopterid, Eumicrotremus. most late-stage demersal em-
bryos resemble the newly hatched larva with respect to all char-
acters (Fig. 1 4G). The morphology of the head, gut, and postanal
body as well as the number of myomeres is used for identifi-
cation within all tish egg groups. A number of specialized char-
acters associated with the embryo are essential for identification
when present, e.g., elongated fin rays— J'rachiplerus (Fig. 14H),
MATARESE AND SANDKNOP: EGG IDENTIFICATION
31
precocious fin development (caudal— exocoetids and Tricho-
don\ pelvic— Trachi mis), and pelvic disc development in some
cyclopterids (Eumicrotremus) (Fig. 14G).
Miscellaneous characters. —The presence of a secondary mem-
brane inside the chorion occurs in some groups, although it is
lacking in most fishes. Sloniias alnvcnter eggs have a double
membrane (Fig. 141). These membranes occur in some of the
more primitive fishes including members of the Anguilliformes,
Clupeiformes, and Salmoniformes. In some species, like the
freshwater cyprinid Abbottina rivularis (Nakamura, 1969), the
secondary membrane is thick and gelatinous. The presence and
size of the micropyle are diagnostic in other fishes, particularly
freshwater demersal eggs (Laale, 1980; Riehl, 1980). Among
freshwater fishes, the cleavage pattern is important for egg iden-
tification. In the more primitive families (Acipenseridae, Poly-
odontidae, Lepisosteidae, and Amiidae), cleavage pattern is typ-
ically semiholoblastic as opposed to the meroblastic pattern seen
in the higher teleosts. Genetic studies have shown differences
in LDH A zymograms to be a useful, diagnostic tool for the
identification of Gadus morhua and Melanogrammus aeglefinus
eggs (Mork et al., 1983).
Ecological and behavioral considerations.- \ number of con-
siderations related to mode of reproduction and collection rather
than the characters of the eggs themselves are essential when
identifying any type offish egg. In identifying demersal eggs one
must consider where they were collected — on rocks, on plants,
in masses, and if parental care is involved. Nest type, nature of
egg deposition, and the presence of guarding parents can all be
essential clues to proper identification. Also, for any egg type
one must note spawning time (season), location depth, and gear
used for collection. In addition, the rearing of unknown eggs to
an identifiable larval stage is useful in species determination as
shown by Stevens and Moser (1982) for the blenny, Hypso-
blennius. Of course, a necessary prerequisite to accurate iden-
tification of eggs is a thorough knowledge of the species present
in any given area and their breeding seasonality.
Summary of Characters
Characters most useful in identification of fish eggs are the
following: ( I ) egg shape— spherical, ellipsoidal, irregular, or oth-
erwise; (2) egg size— fish eggs range in size from 0.5 to 26.0 mm;
(3) oil globules— presence or absence, number, size, color, po-
sition, and pigmentation; (4) yolk — segmented or homogeneous,
nature of segmentation, color, pigmentation, and circulation
pattern; (5) chorion— smooth or ornamented, type of ornamen-
tation, thickness, color, and coatings; (6) perivitelline space-
width; (7) embryonic characters— morphological features, pig-
ment patterns, and special structures; (8) miscellaneous char-
acters—inner or secondary membrane (presence or absence, lo-
cation), cleavage pattern, micropyle (size), and biochemical
analysis; and (9) ecological and behavioral considerations— col-
lection (gear, location, season, etc.), and mode of reproduction
(nests, parental care, etc.).
(A. CM.) National Marine Fisheries Service, Northwest
AND Alaska Fisheries Center, 2725 Montlake Boule-
vard East, Seattle, Washington 98112; (E.M.S.)
Southwest Fisheries Center, P.O. Box 271, La Jolla,
California 92038.
Identification of Larvae
H. POWLES AND D. F. Markle
MINOR errors in identification of larval fishes can lead to
major misinterpretations of ecological and taxonomic
phenomena. Fish identification and taxonomy are largely based
on adult characteristics and since these develop during the larval
period, new characters must be discovered and validated in
order to identify larval fishes. Usually larvae possess fewer char-
acters than adults and are more fragile. Identification can, there-
fore, be difficult and, frequently, must be based on a combi-
nation of character states.
Since larval anatomy is by its nature dynamic (a given spec-
imen being a snapshot of the process linking embryos to adults),
developmental series are essential to identification. Three dif-
ferent approaches are used to identify larvae, the first two of
which arc based on developmental series: I) to raise eggs and
larvae from fertilized eggs of known parents; 2) to work back-
wards from the adult utilizing characters common to succes-
sively earlier ontogenetic stages; and 3) to extrapolate from pre-
vious results obtained by (1) or (2) to synthesize generic or
familial diagnoses and identify by process of elimination or
limited corroboration (Ahlstrom in Berry and Richards, 1973;
Leiby, 1981).
There are pitfalls in all approaches. Laboratory-reared larvae
are frequently more heavily pigmented than wild-caught spec-
imens and may show greater meristic variation (Lau and Shaf-
land, 1982). Laboratory rearing may be financially and logis-
tically difficult or impossible for fishes of interest. Ontogenetic
transformations arc based on associations of adult diagnostic
characters with characters that persist in progressively earlier
ontogenetic stages. This method requires careful attention to
methodology, as well as good ontogenetic series which are not
always available. Purely descriptive accounts of larval series
(laboratory-reared or reconstructed) may not be useful for iden-
tification purposes if no diagnostic characters that will distin-
guish sympatric congeners and/or similar-looking forms are pre-
32
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
sented. Novel sorts of characters or ways of manipulating data
are sometimes needed to identify larvae and the data required
may not be retrievable from "standard" descriptive accounts.
Synthesis and elimination is the normal procedure used by tax-
onomists to identify adult fishes. It has been called the "look-
alike" system when applied to larval fishes (Leiby, 1981). It is
basically a simple procedure but the pitfalls are numerous and
subtle. As with some early adult fish taxonomy, premature syn-
thesis may often be based on the wrong characters (e.g. con-
vergent characters) and lead to spurious identifications.
General references on larval fish identification include Berry
and Richards (1973), Ahlstrom and Moser (1976) and Moser
(1981). Some recent works which provide exposure to a wide
range of larval forms and literature are Ahlstrom and Moser
(1981) and Fahay (1983) for marine taxa, and Auer (1982) and
Balon (1975a, 1981a) for freshwater taxa.
The purpose of the following is to describe the tools— pref-
erably sharpened, polished and comfortable to use— which should
be at hand when the ichthyologist sits down to identify larval
fishes. Our emphasis is on three main factors: 1 ) the larval fish —
its anatomy, ontogeny, and phyletic relationships; 2) the study
area— its ecology and zoogeography and 3) the investigator—
his experience, knowledge and ingenuity.
Systematics, Ontogeny and Anatomy
Perhaps the most important type of character for identifica-
tion of larvae is meristic, as counts usually do not increase or
decrease once established. All meristic characters can be im-
portant, but vertebra/myomere counts and fin element counts
are of particular value. Meristic variables are useful at different
taxonomic levels, e.g., principal caudal fin ray and pelvic fin
element counts at the family or order level, median fin elements
at the genus/species level, pectoral fin ray counts at the species
level. Frequency distributions of meristic counts are extremely
important (particularly when it is uncertain whether develop-
ment of a character is complete) but often are not given in
published literature. Some important characters may not be
included in published studies (e.g., pectoral fin rays, procurrent
caudal rays). Differences in methodology and variable attention
to detail may also affect the quality of published meristic data.
Thus, published studies must be treated with caution and one
must be prepared to collect and compile one's own information
when opportunities arise. Despite potential problems with pub-
lished works, these are the obvious place to start with compi-
lations. Few "regional" meristic publications as exemplified by
Miller and Jorgensen (1973) exist, but many publications on
larval fishes include extensive tabulations of meristic infor-
mation.
Various ways exist for facilitating use of meristic compila-
tions. A simple taxonomic listing (e.g.. Miller and Jorgensen,
1973) can be time-consuming to use, while a "gazetteer" format,
with species arrayed in order of counts (e.g., Fahay, 1983) may
be more practical. X-Y plots of two meristic variables (e.g..
Berry, 1959b) can include frequency distributions and be very
useful for separating closely-related forms.
A second suite of characters of broad use is specialized larval
characters which may characterize whole groups. These include
but are not limited to: characteristic shapes (e.g., Anguilli-
formes/Elopiformes, Pleuronectiformes), spination (Acanthur-
idae, Holocentridae), fin development patterns (argentinoids),
fin element development (Pleuronectiformes, epinepheline Ser-
ranidae), fin placement (pelvic fin placement in Pleuronecti-
formes), eye shape (myctophid subfamilies, salmoniform
groups), and phoiophore development pattern (Gonostomati-
dae). The elucidation of such characters is a focus of this volume,
and reference should be made to specific chapters for further
detail. The important point is that a broad knowledge of larval
fishes is frequently necessary for accurate, efficient identification
of larvae.
Finally, identification of larvae depends on a suite of dynamic
characters (pigmentation, body form, spination, fin develop-
ment pattern, etc.), which may change rapidly and differentially
over a small size range. Generally, a combination of such char-
acters is required for accurate identification; this is particularly
true in early stages. These characters can vary extensively, even
within a species, due to regional differences; method, time or
area of collection; preservation method or duration. Develop-
mental changes can be extremely rapid (e.g., changes in mela-
nophore distribution from some yolk-sac to post-yolk-sac lar-
vae). Again, no extensive treatment of these characters is possible
here, but the important point is that detailed, disciplined ob-
servations of larvae are essential for accurate identification.
The importance of osteological characters for larval identi-
fication is increasingly recognized (Dunn, this volume). Use of
these depends on clearing and staining techniques (PotthofT, this
volume) or X-ray techniques (Tucker and Laroche, this vol-
ume). As with meristics, osteological characters may be useful
at different taxonomic levels. Caudal osteology has been widely
used because of its early development and relative simplicity,
but cranial osteology and pterygiophore patterns are also useful.
Recent application of cartilage-staining techniques has permit-
ted use of cartilaginous structures in identifying larvae (e.g.,
Fritzsche and Johnson, 1980). Other internal characters such as
gut shape (Ahlstrom and Moser, 1976; Govoni, 1980) may also
be useful.
Keys have not generally been used in larval fish identification
because of the dynamic nature of characters (a separate key
would be required for each size class or development stage) and
because of "incompleteness" of information (i.e., it has usually
been impossible to completely cover a defined region or sys-
tematic group with a key). Generally, much more information
is required to identify a larva than an adult, and summarizing
this in a key has been impractical (the information-organizing
capacity of computers may eventually help to permit this). Ex-
ceptions, such as Bertelsen's (1951) key to larval Ceratioidea,
Johnson's ( 1 974b) key to genera of larval scopelarchids, and the
key of Bertelsen et al. (1976) to notosudids do exist.
Because of the complexity of identification of larvae, a wide
ichthyological background is important. A good knowledge of
fish anatomy is essential, particularly when (as often occurs)
damaged specimens must be identified. Published descnptions
exist, for example, which interpret broken branchiostegal rays
as jugular pelvic fin rays. A general knowledge of suspected
phylogenies and inter-relationships (e.g.. Greenwood et al., 1966;
Nelson, 1976) is essential if attempting to identify by synthesis
or elimination. This should at least cover those groups to be
expected in a given area, but wider knowledge is desirable, par-
ticularly in the marine environment where exotic larvae may
be transported great distances (e.g., Markle et al., 1 980). Finally,
thorough familiarity with the ontogenetic continuum is neces-
sary to place unknown specimens in perspective. Absorption of
the yolk sac, flexion of the notochord in the caudal region,
development of median fins, and transformation from larval to
POWLES AND MARKLE: LARVAL IDENTIFICATION
33
juvenile stages (as defined by completion of fin element devel-
opment, development of scales, etc.) are major events in fish
development which have been used by various authors to define
stages (e.g., Ahlstrom, 1968; Snyder, 1976).
Ecological Considerations
There are two basic ecological or zoogeographic consider-
ations when identifying larvae: the expected composition of the
larval ichthyofauna of the study area and the potential for influx
from "upstream" areas.
Thorough knowledge of the adult ichthyofauna of the study
area is essential in order to know what larvae may occur; thus,
the most complete possible list of adult species is required.
Literature may be incomplete or erroneous, so this list should
be based on unpublished or personal observations as well as on
standard faunal works or other literature. For ease of use, the
list should be organized by systematic groups (e.g.. Greenwood
et al., 1966; Nelson, 1976).
In addition to knowledge of the adult ichthyofauna, knowl-
edge of spawning seasons is central to prediction of the larval
fish composition. As with meristic or anatomical information,
published information may be incomplete so that personal col-
lections and unpublished information may be important. Al-
though capture location and season can be important in elim-
inating some species from consideration, caution is essential
here as with other "elimination" methods.
Since most marine fishes have planktonic eggs and/or larvae
and many have a prolonged planktonic life the basic hydrog-
raphy of a study area must be understood. A "downstream"
study area is potentially vulnerable to an influx of larvae from
"upstream" spawning. In addition, the direction of "streams"
can differ at different depths of the water column so the influx
may come from more than one direction. On the shelf oR"Nova
Scotia the general circulation is from the northeast but there is
a strong influence from the Gulf Stream, both from eddies and
mixing which produces Slope Water. Thus, for some species,
the "downstream" effect comes from the northeast while for
tropical and oceanic species it comes from the southeast.
Knowledge of an area's fish communities may help in inferrmg
which larvae may occur together— for example, an unknown
specimen taken together with larvae from a coastal community
is probably not a mesopelagic species. Again, however, such
inferences should be considered critically.
One sort of ecological observation may be misleading— al-
though spawnmg biomass may be calculated from egg and larval
abundance for some species, the relative apparent abundance
of adults is not always in proportion to the relative abundance
of planktonic larvae. Cryptic species may appear rare in collec-
tions of adults but larvae may be extremely abundant (e.g.,
Gobiidae in tropical and subtropical waters) while species which
appear extremely abundant as adults may be rare as planktonic
larvae (e.g., the clupeid Jenkmsia lamprotaenia in the Carib-
bean, Powles, 1977).
Some General Considerations
Like larval development, identification of larvae is a dynamic
process— the cumulative knowledge of the student is the key to
accurate identification. The complexity of larval identification
requires that a wealth of information be applied to the task, and
for this reason some degree of specialization in identification of
larvae is required for all but the simplest identification prob-
lems. There are many examples of superficially similar but sys-
tematically very different larvae, and most students, including
the authors, have experienced embarrassment at an uncritical
identification. Identification of larvae is frequently comparative,
by elimination, so that wide knowledge of larval fishes as well
as caution are necessary.
The student must have information of the kinds identified
above. Organization and ingenuity are required in order to keep
this information usable — card files, looseleaf binders, drawings
and sketches, and well-curated reference series should be de-
veloped or readily available.
Finally, although many beginning students are hesitant to
draw, sketching and drawing (freehand, on squared paper, or
with camera lucida) is one of the best ways to "see" and un-
derstand larval anatomy. The process is painstaking and often
frustrating in the early stages, but will pay off in the long term
with increased understanding.
(H.P.) Fisheries and Oceans, P.O. Box 15500, Quebec GIK
7Y7, Canada; (D.F.M.) Huntsman Marine Laboratory,
Brandy Cove, St. Andrews, New Brunswick, EGG 2X0
Canada.
Illustrating Fish Eggs and Larvae
B. Y. SuMiDA, B. B. Washington and W. A. Laroche
SCIENTIFIC illustrations of fish eggs and larvae are an in-
dispensible component of any descriptive work, providing
a visual reference of form and structure which is not possible
to express by written descnptions and measurements alone.
Illustrations facilitate identification by emphasizing distinctive
but often subtle morphological characters and allow for com-
panson of features at difl^erent developmental stages and with
morphologically similar taxa. These qualities make illustrations
the preferred and most frequently used aid for taxonomic iden-
tification of fish eggs and larvae.
The broad range of morphological diversity found among
larval fishes requires flexibility in technique and style to produce
eflTective illustrations, but the criteria of accuracy, clarity, and
consistency of style should be met. The basic concept behind
illustrating a fish larva involves accurately representing a three-
dimensional, somewhat transparent organism on a two-dimen-
34
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
sional sheet while emphasizing characters which are most useful
in identifying the actual larva from the drawing. Such characters
include the fins, pigmentation patterns, and details of the head
such as the jaws, spines and eyes. Internal structures such as
myomeres, the gut, cleithrum, and posterior end of the noto-
chord may also be emphasized but without masking important
external characters. Details of other internal structures as well
as shading or stippling for contrast are best excluded or de-
emphasized to maintain clarity. Pigmentation is important in
identification of most larvae and should be depicted clearly.
External melanophores can be drawn with a fine-tipped pen as
realistically as possible. Internal pigmentation can be effectively
represented by using light stippling with a smaller sized pen-
point. Care must be taken to avoid confusion of internal struc-
tures with pigmentation.
Specimens selected for illustration should ideally be those of
the best condition available and representative of the particular
developmental stage in both pigmentation pattern and mor-
phology. The number of specimens to be illustrated is deter-
mined by the nature and objective of the publication, the amount
of material available in various size groups, and the degree of
morphological and pigmentation change undergone by the par-
ticular species during ontogeny. Specimens from described series
should be archived in a museum collection for proper care and
future reference after completion of the illustrations, and catalog
numbers should be published.
The detailed drawing begins with an accurate body outline
showing the proper body proportions and position of fins and
critical pigment spots. This is most easily achieved by drawing
in light or blue pencil from a camera lucida-equipped micro-
scope. Other methods include drawing from a projection of a
slide transparency of the specimen or tracing a photograph. By
convention the lateral view of the larva is drawn, with the head
to the left. The exception to this is made with right-eyed pleu-
ronectiforms. In some instances a dorsal or ventral view is also
necessary to clarify a pigment pattern or laterally projecting
morphological structures. If sketching through a camera lucida,
it is helpful to use a magnification which allows the entire spec-
imen to be in the field of vision as long as important details
remain visible. Any resulting distortions at the periphery of the
field can be compensated for by differentially focusing the mi-
croscope on the particular region involved while carefully pen-
cilling along the image, then reconstructing a smooth line where
disjointed lines meet. Problems involving specimens that are
too large or too small can often be overcome by using lens
adapters or eyepieces of lower or higher magnification. Large
specimens may require being drawn in sections which are later
pieced together. This original sketch should be made large enough
to clearly indicate fine details such as the full complement of
fin rays, but not excessively so with the result of producing lines
which bleed in the final reduction for publication. Related to
this is the use of appropriate sizes of pen points which produce
lines fine enough to draw minute details yet not be lost in re-
production. Therefore, in determining the original size of each
drawing, thought should be given to the desired reduction ratio
as well as the number of illustrations comprising each plate. An
opaque projector is most useful for obtaining a specific size for
the final drawing from the initial sketch, but photocopy reduc-
tions also work well. With this final pencilled sketch, the illus-
trator can work with the larva under a microscope as a reference
to complete details of the drawing before attempting to ink it.
A light table can be helpful when tracing or inking over a rough
pencilled sketch. The illustrator should always have a set of
meristics of the specimen being drawn and an understanding of
the important characters to be emphasized. A thorough inspec-
tion for accuracy is essential to insure agreement between il-
lustrations and descriptive text, especially concerning pigmen-
tation and meristic elements with size and stage of development.
Ideally exact counts and measurements can be obtained directly
from the illustration, allowing easy identification of the larva.
Illustrations are often designed for comparison of features at
different stages of development or for comparison of similar
features which occur among different taxa. Special care should
be taken to represent similar features in a consistent style from
illustration to illustration. For example, a partially ossified fin
ray element, an ossified fin ray, and a fin spine may each be
depicted in a consistent but slightly different manner so that the
illustration not only shows the number and position of fin ele-
ments but also the type of element and its relative stage of
development.
Literature dealing with larval fishes contains a broad array of
illustrative styles, techniques, and quality. Many of these are of
limited use since they fail to meet the criteria discussed above.
Photographs frequently yield unsatisfactory results due to dif-
ficulties in focusing on small, transparent organisms so that all
body parts appear equally sharp, and they preclude emphasizing
inconspicuous but important features for identification. Color
illustrations in a variety of media, although potentially valuable,
particularly for xanthophores, are limited due to prohibitive
publication costs, poor reproducibility, and the absence of a
long-lasting color preservative. Half-tone illustrations (see Ahl-
strom, 1965) are effective but difficult to reproduce. These latter
two techniques may become more practical with advances in
photocopy technology. The preferred technique in widespread
use consists of pen and ink drawings done in black India ink.
Various styles of illustrations of diverse groups of larvae are
represented in Moser (1981) and in this volume which serves
as a useful overview. Poul Winther, George Mattson, and other
artists (Ahlstrom and Ball, 1954; Ahlstrom and Counts, 1955;
Bertelsen and Marshall, 1956; Ege, 1953, 1957, and 1958; Grey,
1955b; Moser, Ahlstrom and Sandknop, 1977; Moser and Ahl-
strom, 1970; Tuning, 1961; Richardson and Washington, 1980)
have been instrumental in establishing a fine style of pen and
ink drawings which we emulate and have found most effective
in its applicability to larval fish identification. We maintain a
degree of flexibility in technique and style which varies with the
taxonomic group under consideration but falls within the gen-
eral framework discussed above.
Illustrating a fish egg poses a more difficult problem than
illustrating a fish larva and will be limited to a brief discussion.
Encapsulation by the chorion necessitates representing the three-
dimensional quality of the egg in the drawing while showing
important morphological and pigmentation characters of inter-
nal structures (Ahlstrom and Moser, 1980; Matarese and Sand-
knop, this volume) with as much clarity as possible. Difficulties
arise due to the superimposing of these characters from a two-
dimensional perspective, particularly when the chorion is or-
namented, when an oil globule(s) is present, and when the de-
veloping embryo is fully coiled.
In spite of the more complex structural representation re-
quired, the same criteria of accuracy, clarity and consistency of
style apply to egg illustrations. The relative proportions of the
egg size to the size of the embryo, oil globule(s), and width of
perivitelline space, the number of myomeres, and length of gut
SUMIDA ET AL.: ILLUSTRATING
35
need to be accurately drawn. An effective balance between show-
ing important characters for identification and three-dimen-
sional reahsm of the egg is required to maintain clarity. Several
illustrations of the egg at different stages of development and
from different perspectives are helpful in demonstrating key
characters such as embryonic pigmentation, myomeres, and po-
sition of the oil globule(s) in the yolksac. Adherence to a con-
sistent illustrative style is primarily critical for a developmental
series of eggs. As with fish larvae, pen and ink drawings provide
the most practical technique for illustrating fish eggs, but the
specific style of illustrating and details shown depend upon the
character of the egg and its stage of development. Many kinds
of illustrative styles and techniques are found in the literature
(see Ahlstrom and Moser, 1980 and references cited therein)
and examination of these is most helpful in effectively illus-
trating a particular type of fish egg.
(B.Y.S.) National Marine Fisheries Service, 8604 La Jolla
Shores Drive, La Jolla, California 92038; (B.W.) Gulf
Coast Research Laboratory, East Beach Drive, Ocean
Springs, Mississippi 39564; (W.L.) Department of
Fisheries, Humboldt State University, Arcata, Cal-
ifornia 95521.
Clearing and Staining Techniques
T. POTTHOFF
THE clearing of tissues and the staining of cartilage and bone
are indispensable in the study of larval and juvenile fishes.
At the National Marine Fisheries Service Miami Laboratory
modifications of the clearing and differential cartilage-bone
staining technique proposed by Simons and Van Horn (1971)
and Dingerkus and Uhler (1977) are used. The modifications
are in part based upon an unpublished manuscript by W. R.
Taylor and G. C. Van Dyke from the National Museum of
Natural History, Washington, D.C. A wide size range of fish
from 3 mm NL to larger than 500 mm SL can be cleared and
stained. The technique works well for all sizes, but adjustments
in the various solution soaking times are made dependent on
fish size (Table 5).
Method
F/.Ya/ZoA!. —Specimens are fixed in 1 0-15% marble chip buffered
formalin. Samples previously fixed in formalin of lower than
10-15% concentration and specimens presently in alcohol or
fixed in alcohol should be refixed in 10-15% formalin for
best results. Eighty to 90% of all larvae of different perciform
families fixed in alcohol totally disarticulated during clearing
and staining. In juvenile and adult fish > 100 mm SL the flesh
is routinely removed from the left side before or after fixation.
Dehydration— This is an important step, because even small
amounts of water interfere with the staining of cartilage. Place
specimen from the formalin into solution of 50 parts of 95%
cthanol and 50 parts distilled water. Do not wash or soak spec-
imens with water during transfer from formalin to alcohol.
After one day for larvae < 20 mm SL and two days for specimens
20-80 mm SL and three to five days for specimens >80 mm
SL transfer from 50% ethanol into absolute ( 100% or 200 prooO
ethyl alcohol. If absolute ethanol is not available, 190 proof or
95% ethanol can be substituted for the absolute, although stain-
ing of cartilage will not be as intense. A second change of ab-
solute alcohol is desirable in larger than 20 mm SL specimens.
Leave larvae <20 mm SL for one day in the absolute alcohol
and juveniles 20-80 mm SL for 2 days. Adult and juvenile fish
80-200 mm SL should be kept in absolute ethanol for 3 days
and fish >200 mm SL should be soaked for one week. An
intermediate absolute alcohol change should be given to all
specimens with longer than one day soaking time.
Cartilage staining. — This is accomplished by placing specimens
in an acidified alcohol solution of the alcian blue stain. For best
results 70 parts of absolute alcohol should be mixed with 30
parts of acetic acid 99% glacial. To every 100 ml of acidified
alcohol 20 mg of alcian blue powder should be added. The above
solution should be used on larvae and juveniles from 3 mm NL
to 80 mm SL. For larger fish, a staining solution of 60 parts
absolute alcohol and 40 parts of acid with 30 mg of alcian blue
for every 100 ml of acidified alcohol should be used. Fish larvae
and juveniles <80 mm SL should be left in the alcian staining
solution no longer than 24 hours. Larger juveniles and adults
should be stained no longer than 36 hours. Specimens >500
mm SL can remain 48 hours in the alcian staining solution.
After the specified time in the alcian solution the stain is per-
manently fixed in the cartilage and cannot be removed with any
chemicals used in the clearing and staining process. Staining
solution can be used twice for staining larvae but should be
discarded after staining a juvenile or adult fish.
Neutralization. — This process raises the pH within the specimen
thus allowing proper subsequent bleaching. The higher pH pre-
vents further calcium loss from the bones for better alizarin red
stain. To neutralize the specimen remove it directly from the
alcian staining solution and place it in a saturated sodium borate
solution for 12 hours for specimens <80 mm SL and for 48
hours for larger specimens. For the juveniles and adults that
soak for 48 hours, change the sodium borate solution once.
Bleaching (an optional .s/cpA — Larvae with little pigment on
their body (e.g., Scombridae) should not be bleached. Larvae
covered with pigment (e.g., Istiophoridae) and all juveniles and
adults must be bleached. Prepare bleaching solution by mixing
36
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 5. Method of Clearing and Staining Cartilage and Bone in Larvae, Juvenile and Adult Fish.
Length in mm, NL or SL
Steps
10
20 .10
40
50
60 70 80 90
100
200 .100 400 500
>500
Fixation:
10-15% formalin
marble chip buffered.
--►h -
— 5 days, flesh removed—
on left side
...-►
Dehydration:
1. 50% distilled H,0,
50%of95%ethanol.
2. Absolute ethanol
(95% ethanol may
be substituted).
-1 day ►[-
-1 day ►(-
h
2 days •
2 days
>h-
3 days ■
►h 3 days
-one intermediate change
-►h •
-5 days-
-7 days-
-►
■-►
Staining cartilage:
100 ml solution:
A. 70 ml absolute ethanol,
30 ml acetic acid,
20 mg alcian blue.
100 ml solution:
B. 60 ml absolute ethanol,
40 ml acetic acid,
30 mg alcian blue.
— I day
-Solution A-
-►h I'/idays ►h"2 days->-
-►H Solution B ►
Neutralization:
saturated sodium
borate solution.
'/2 day -
-►I--
I-
2 days
-one intermediate change -
-►
-►
Bleaching:
pigmented
specimens only.
100 ml solution:
15 ml 3% H,0„
85 ml 1% KOH.
-20 min.
-►!-■
-40 min.
-► h - 1 hour ► I 1 1/2 hours -
Trypsin digestion:
100 ml solution:
35 ml saturated sodium
borate. 65 ml distilled
H,0. trypsin powder.
-Keep in solution until 60% clear, change to fresh solution every 10 days-
Staining bone:
1% KOH solution with
alizarin red stain.
I day -
-►[-■
2 days •
-►h
-4 days
Destaining:
100 ml solution:
35 ml saturated sodium
borate, 65 ml distilled
HjO, trypsin powder.
-2 days — ►!- ■
Change to fresh solution every 10 days until solution remains -
unstained and specimen is clear
Preservation:
30% glycerin and
70% of 1% KOH.
60% of glycerin
and 40% of I % KOH.
1 00% glycerin with thymol
as final preservative*.
1 week — ►!- - -2 weeks- — ►!- 4 weeks-
* Direct sunlight and 100% glyceiine help to clear and destain difficult specimens.
15 parts of 3% hydrogen peroxide solution with 85 parts of
1% potassium hydroxide solution. Bleach larvae and small ju-
veniles up to 80 mm SL for 20 to 40 minutes depending on
size. Larger juvenile fish and adults may be bleached 1 to 1 Vi
hours.
Trypsin digestion and alizarin red staining. — The clearing and
alizarin staining process has been well described by Taylor ( 1 967)
and need not be repeated here. Simply continue after bleaching
with the Trypsin digestion, which are Taylor's steps 4 and 5.
We saw no need in modifying Taylor's method.
Removal of semitransparent tissue. ~^\\ex\ studying cleared and
stained material of large fish, the structures studied (caudal com-
plex, pectoral fin supports, pterygiophores, vertebral column,
etc.) may have to be dissected out and adhering tissue removed.
This can be accomplished by time consuming picking with
tweezers or by placing the material in a two-phase phenol so-
POTTHOFF: CLEARING AND STAINING
37
lution with the addition of heat (Miller and Van Landingham,
1969). With this method the bones are not disarticulated, but
some bone distortion was experienced.
Variables affecting results.— The results of the clearing and
staining procedure are not always satisfactory because of known
and unknown variables. Results can never be predicted with
certainty. The known variables are: ( 1 ) Time and ambient tem-
perature the organism is subjected to between death and fixation.
The longer an organism remains unpreserved after death and
the higher the temperature, the less the tissues will clear. For
best results, specimens should be killed in the fixative, or if that
is not possible, they should be kept cool or frozen before fixation.
(2) Effect of fixative and preservative. Marble chip buffered
formalin is a good fixative for larval fish if specimens are re-
moved from it after 24 hours. Buffered formalin as a preser-
vative destroys first the stain uptake in cartilage. Bone decalcifies
as buffered formalin becomes acid over a longer time period
and decalcified bone will not stain. Therefore, it is best to fix
specimens in 10% formalin and then to preserve them in 70-
95% ethanol. Specimens fixed and preserved in ethanol should
be re-fixed in formalin before clearing and staining. (3) Time in
a preservative. The longer a specimen has been preserved, the
less predictable the clearing and staining outcome will be. Some
fish larvae from the Dana collection in the 1920's were cleared
and stained. The results were startling for both Formalin and
alcohol preserved material because some specimens cleared and
stained well, but most were unfit for study.
Other vanables which affect the results of clearing and staining
exist, but are not understood. No matter how carefully one
adheres to the procedures, the clearing and staining results are
not predictable.
Interpretation of results. — Frequently specimens will remain
opaque and overstain with alcian or alizarin for unknown rea-
sons. This makes viewing of cartilage and bone structure diflicult
or impossible. Such specimens can be used for study of fin ray
development and for fin ray counts.
Cartilage or bone does not always stain but can be made
visible in cleared preparations by changing light conditions at
the microscope and manipulating the substage mirror. Cartilage
appears reticulated in structure whereas bone is structurally clear
and hyaline.
Erroneous conclusions can be made if one solely relies on
color to determine cartilage and bone. In general, cartilage will
appear blue and bone red, but often alcian blue is taken up by
bones and rarely alizarin red by cartilage. For instance, devel-
oping fin rays often appear blue.
Generally larger developed cartilage structures will stain bet-
ter than small developing ones. Thus, in the same specimen one
may find brightly blue stained cartilage, pale blue cartilage, and
cartilage with no stain at all. Therefore, special care is indicated
when viewing newly developed cartilage.
The ossification onset in cartilage is difficult to determine. A
thin layer of bone forming all around the cartilage can be de-
tected by examining the outer edges of the cartilage structure:
a shiny hyaline line forms there, probably only a cell layer thick.
Investigators are often discouraged by clearing and staining
results, particularly when their sample is small. In a larval de-
velopmental series I usually clear and stain 200 to 400 speci-
mens, and I am able to study each aspect and area of devel-
opment that I wish to examine because of the large sample size
at hand. For example, in a specimen in which the pectoral fin
support area is unclear and stained poorly the caudal area may
be clear and stained well. Thus, this specimen is utilized only
for caudal development, whereas in another specimen the pec-
toral area may be clearer and better stained. Thus, with a large
sample size, the uncertainties and vagaries of the clearing and
staining procedure are overcome.
Application of clearing and staining.— Cleanng and staining is
helpful in identification offish larvae when external characters
are inadequate. It also aids systematic and phylogenetic studies
of larvae to adult fishes. This subject has been discussed in detail
by Dunn (1983b).
National Marine Fisheries Service, Southeast Fisheries
Center, Miami Laboratory, 75 Virginia Beach Drive,
Miami, Florida 33149.
Radiographic Techniques in Studies of Young Fishes
J. W. Tucker, Jr. and J. L. Laroche
RADIOGRAPHY is useful for obtaining skeletal informa-
tion in studies of fish taxonomy and morphology. Al-
though clearing and staining provides more detail, radiography
has other advantages. It produces an easily stored, long-term
record of the skeleton and does not permanently alter the con-
dition of the specimen. In many cases, counts can be obtained
more accurately from radiographs than from the specimens
themselves. If an x-ray unit and darkroom are available, ra-
diography is usually faster and easier than clearing and staining.
The time saved may be of value in studies of population vari-
ation, in which many specimens must be examined. Radiog-
raphy has also been used to monitor decalcification of larvae
stored in formalin (Tucker and Chester, in press), and has been
suggested for use in toxicological studies to check large numbers
38
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
of larvae for skeletal deformities. The consensus among ichthy-
ologists who have used both techniques is that, although clearing
and staining methods provide the detail necessary for describing
developmental osteology, radiography is a simple and quick way
of obtaining counts from large numbers of specimens.
Hard (shortwave) x-rays have been used to form shadow pic-
tures, or radiographs, of large, well-ossified fish for almost four
decades (Goshne, 1948; Bartlett and Haedrich, 1966), but the
use of soft (longwave) x-rays for small specimens is relatively
new. Although first suggested by Bonham and Baylifr( 1953) and
used by Watson and Mather (1961 unpubl. manusc), useful
techniques for larval radiography have only recently been de-
scribed (Miller and Tucker, 1979). Potential larval fish radiog-
raphers should consult Miller and Tucker's paper for method-
ological details and Quinn and Sigl ( 1 980) for basic radiographic
principles. Although specimen fragility determines the mini-
mum size of larvae that can be x-rayed, sensitivity of the tech-
nique, which depends to a large degree on spectral characteristics
of the radiation, determines the amount of detail present in the
finished radiograph. This section, therefore, reviews the prin-
ciples and current methods useful for maximizing detail in ra-
diographs of fish larvae.
Radiographic sensitivity refers to the clarity of details in the
radiographic image and depends on a combination of two fac-
tors, definition and radiographic contrast. Definition is sharp-
ness of the image. Radiographic contrast refers to the density
(darkness) range of the image and depends on two factors, sub-
ject contrast and film contrast. Subject contrast refers to the
ratio of radiation intensities that pass through different parts of
the specimen. Film contrast refers to the ratio of densities in
parts of the film that have received different degrees of exposure.
In larval fish work, radiographic sensitivity can be improved
by several means. Definition can be improved by using the
longest possible radiation wavelengths, by using the finest grained
film available, and by minimizing geometric production of over-
lapping shadows at tissue discontinuities in the specimen. Ab-
sorption by x-rays of a given wavelength depends mostly on the
atomic numbers of components in the x-rayed material, and to
a lesser degree on thickness and density of the material. Larval
skeletons, which are thin, poorly calcified, and of relatively uni-
form composition and thickness, do not contrast radiographi-
cally with the rest of the body as much as in older fish. High
contrast techniques should, therefore, be employed. Subject con-
trast can be increased by increasing wavelengths and by de-
creasing the thickness of non-skeletal tissue by dehydrating the
specimen. Film contrast can be increased by using a high con-
trast film and by increasing development time; however, over-
development will also increase graininess and reduce definition,
and probably should be avoided.
The longwave (soft) end of the x-ray spectrum is the portion
most useful for x-raying small fish, because this low energy
radiation does not pass through materials as easily as that at
the shortwave (hard) end. Decreasing the tube voltage (kv) caus-
es a shift of the emitted spectrum toward longer wavelengths.
Resultant elimination of some of the hard radiation contributes
to better subject contrast and improves definition by reducing
clumping of silver grains in the film emulsion (graininess). The
x-ray unit should be equipped with a thin beryllium window,
which allows passage of soft rays. A 25 mil (0.63 mm) window
allows work at a kv of 20; a 10 mil (0.25 mm) window extends
capabilities to about 8 kv (Joseph Fowler, Hewlett Packard, pers.
comm.). However, the lower practical limit for fish larvae may
be governed by restrictions on exposure time, rather than kv
limitations.
Another relevant factor is the source-to-specimen distance,
to which image definition is directly related. Increasing the source-
to-specimen distance improves definition by minimizing en-
largement and distortion. Practical limits are set by air atten-
uation, loss of radiation intensity (roughly as the square of the
ratio of the distances), and dimensions of the x-ray unit. Geo-
metric unsharpness is the maximum width of the zone of over-
lapping shadows that are caused by a non-point source. This
factor can be calculated to determine the minimum source to
specimen distance that can be tolerated. Use of the minimum
distance will permit the shortest possible exposure time and
reduce relative attenuation of soft rays, thus contributing to
subject contrast. The formula for geometric unsharpness, Ug
(Quinn and Sigl, 1980) is:
U„
D,
in which F is the radiation source size. Do is the source-to-
specimen distance, and t is the specimen to film distance (max-
imum specimen thickness). For F = 0.5 mm, D,, = 460 mm,
and t = 1 mm, U^ is 0.00 1 mm. This level of unsharpness would
not be visible without magnification and could be tolerated at
moderate magnification depending on the requirements of the
investigator. To ensure that geometric unsharpness is not large
enough to affect quality of radiographs, it should be calculated
for the set of factors relevant to each operation, keeping in mind
the level of magnification to be used. With most modem x-ray
units, a distance of 46 cm or less can be used.
Because air attenuates soft rays more than hard, elimination
of air between the x-ray source and specimen allows a greater
proportion of soft radiation to reach the specimen. Decreasing
the source to specimen distance helps some, but also increases
geometric unsharpness, unless the source is very small. A vac-
uum would be ideal but is impractical. Replacement of the air
in a cabinet unit with helium allows the use of lower kv with
reasonably short exposure times and provides an increase in
subject contrast. Helium can be conserved and reused if it is
placed in a small volume plastic cylinder that has its ends sealed
with dry-cleaning plastic.
Before a specimen is x-rayed it should be dehydrated as much
as can be tolerated to increase the signal (skeleton) to noise
(non-skeleton) ratio. For best results, the specimen should be
placed in 50-75% ethyl alcohol for a short period, maybe 30-
60 min, depending on size. Then the specimen should be placed
on the film holder, blotted to remove surface liquid and bubbles,
and quickly x-rayed and returned to a container of liquid before
desiccation damage occurs.
The specimen should be placed as close as possible to the film
emulsion. This can be accomplished without wetting the film
by sandwiching it between two thin sheets of black polyethylene.
Details for construction of a convenient film holder (cassette)
are presented in Miller and Tucker (1 979). Polyethylene is trans-
parent to soft x-rays and is good cassette material. Vinyl, as well
as wood, paper, and any metal are relatively opaque to soft
x-rays, and vinyl or metal make good labels.
Single coated Type R (now Type XAR) film has provided the
best quality radiographs of larvae. High resolution plates give
better resolution but are too slow. Type R film is slow relative
to other films but within practical limits. It has ultra-fine grain
TUCKER AND LAROCHE: RADIOGRAPHY
39
Fig. 15. Positive image of radiograph of a southern flounder (Paralichthys tethosligma) larva, 9.7 mm SL, stored m 7% borax buffered seawater
formalin for seven years. Radiographic exposure data: Faxitron Model 43805N; Kodak Type R film; source to film distance. 46 cm; 9 kv; 600
mAs; under helium. Intemegative processing data: radiograph was projected onto 4 in x 5 in professional copy film (Kodak 4125) with an Omega
(4 in X 5 in) Pro Lab Enlarger; exposure was 1 s at f S'/j; film was developed in Kodak HCl 10 (dilution E) for 5 min at 23 C. Print processing
data: a positive pnnt was made on Kodak Polycontrast Rapid 11 RCF paper using a polycontrast no. 3 filter in the Omega enlarger; exposure was
5 s at f 5.6; print was developed in Kodak Ektaflo diluted to simulate Dektol 1:1, at 23 C. (The intemegative and printing procedure was devised
and performed by Tom Smoyer of Harbor Branch Foundation.)
and high contrast. The single emulsion is necessary for avoiding
two images (on both sides of the film). Coarser grained and
lower contrast films will produce inferior radiographs.
Exposures should not be longer than about 5 min, and for
many specimens 5 min is too long. Larvae will quickly desiccate,
and even if not damaged, may shrink and cause blurred images.
Specimen damage or image blurring will determine the mini-
mum size of larvae that can be x-rayed. Specimens can be pro-
tected by an overlying sheet of dry-cleaning plastic if care is
taken to remove bubbles. During exposure, unneeded portions
of the film can be protected for later use with lead vinyl masks.
The manufacturers' instructions for mixing chemicals and
processing films should be followed as closely as possible. Fre-
quent agitation of the film while it is developing, rinsing, and
fixing is important to ensure uniformity of chemical reactions.
Both undeveloped and developed films should be stored away
from light, heat, humidity, and chemical fumes (particularly
formalin, alcohol, and hydrogen peroxide). Radiographs are best
observed directly, emulsion side up, with a dissecting or phase
contrast microscope. Printing of radiographs is best done via
an intemegative (Fig. 15). This compresses the tonal range so
that finer detail can be preserved in the print.
The major limitation of the technique is probably inadequate
radiation intensity at low kv. This limit may have been reached
with x-ray units equipped with 10 mil beryllium windows. Sat-
isfactory radiographs of 4-1 5 mm larvae have been made at 8-
10 kv and 300-800 mAs (milliamperes x seconds). Some im-
provement can be expected if the air is replaced with helium;
however, exposure time will eventually become prohibitively
long.
Because machine and specimen characteristics vary, a stan-
dard formula for producing high-quality radiographs cannot be
provided. At least initially, the larval fish radiographer must
proceed by trial and error with the machine and specimens at
hand. As familiarity develops, the results will improve signifi-
cantly. We stress that an accurate and detailed logbook con-
taining specimen and exposure data should be kept, and that
procedures should be standardized.
(J.W.T.) Harbor Branch In.stitiition, Inc., RR l,Box 196-A,
Fort Pierce, Florida 33450; (J.L.L.) Gulf Coast Re-
search Laboratory, East Beach Drive, Ocean Springs,
Mississippi 39564.
Histology
J. J. GOVONI
WHILE contemporary systematists rely upon a broad scope
of biological features to infer relationships among taxa,
the definition and comparison of morphological characters re-
mains one of their most useful tools. The small size and often
altricial development of fish larvae, however, make it difficult
to resolve the morphology of structures other than skeletal ele-
ments. By clarifying tissue composition and by enhancing mor-
phological resolution, histological techniques may aid the sys-
tematist in defining characters at the tissue as well as at the
microanatomical level, thereby providing additional character
states to be examined for synapomorphies and perhaps onto-
genetic precedence. Because of their small size, sections of whole
larvae can be prepared (Fig. 16) and structural relationships of
organ systems examined. Insofar as there is no clear separation
between gross and micro-anatomy beyond the limits of human
visual resolution, histological techniques may otfer yet another
tool useful in phylogenetic analysis.
Techniques
Flvi2;/o«. — Inasmuch as autolysis is rapid in larval tissue (Thei-
lacker, 1978), fixation is difficult (Richards and Dove, 1971).
Specimens reared in the laboratory or specimens taken from
brief plankton tows (O'Connell. 1980) are the most suitable for
histological preparation and study; specimens sorted from field
collections fixed in formalin and seawater will usually yield poor
quality preparations. Neutral buffered (phosphate buffi;rs) for-
malin (see Humason, 1979) enhanced with <4% acrolein (van
der Veer, 1 982) is recommended for rapid and thorough fixation.
Glutaraldehyde (2.5%) is also a useful fixative (Hulet, 1978).
Difference in the osmolality of tissues and ambient water may
distort cells and tissues, especially of marine larvae. Such arti-
facts have not been observed in preparations of clupeiform and
perciform larvae, but may be of concern in the preparation of
anguilliform leptocephali (Hulet, 1978). Forsterand Hong (1958)
and Hulet (1978) provided applicable saline solutions that may
eliminate distortion and enhance staining.
Sectioning and staining. — Sxandsivd animal tissue techniques
(e.g., Humason, 1979)— dehydration, paraffin embedding, and
sectioning— have been used to trace the development of organ
systems (O'Connell, 1981a), as well as to assess the pathology
of starvation in fish larvae (Umeda and Ochiai, 1975; O'Con-
Fig. 16. Sagiual section ot a Leiostomus xanlhurus larva, 4.4 mm notochord length (glycol methacrylate section stained with alkali blue 6B-
neutral red).
Fig. 17. Example comparisons of larval fish tissue and microanatomy. Abbreviations: AM, axial musculature; CS, collagenous supporting
shafts; EP, epidermal cells; M, midgut; MC, mucous cell; NF, nerve fiber. (A) The integumentary epithelium of a Brevoortia patronus larva
showing hyaline plates (arrow), a tissue characteristic of some clupeiform larvae. Note that erosion of the outer layer of epithelium is evident.
(Scale bar = 20 /jm; glycol methacrylate section stained with acid fuchsin — toluidine blue.) (B) The integumentary epithelium of a Leiostomus
xanthurus larva showing lack of hyaline plates in epithelial cells. (Scale bar = 10 iim; glycol methacrylate section stained with alkali blue 6B —
neutral red.) (C) Axial musculature of a Brevoortia patronus larva showing two opposing layers of muscle fibers, a tissue characteristic of clupeiform
larvae. (Scale bar = 50 livn, glycol methacrylate section stained with acid fuchsin — loluidine blue.) (D) Axial musculature of a Leiostomus xanthurus
larva showing muscle fiber layers in parallel alignment, a tissue characteristic of perciform larvae. (Scale bar = 50 iim\ glycol methacrylate section
stained with alkali blue 6B — neutral red.) (E) Cross section of the elongate dorsal ray of an Echiodon dawsoni larva. (Scale bar = 20 ixm: glycol
methacrylate section from Govoni et al., 1984.) (F) Cross section of the elongate dorsal ray of a Bregmaceros atianticus larva. (Scale bar = 15
Min; glycol methacrylate section stained with acid fuchsin — toluidine blue.)
40
GOVONI: HISTOLOGY
41
.:•»' /■
B
.f \
MC
tiSNHBC^-
EP
AM
'^ .- i^.
t
w
•tr\
►'i'^ .
,^
" f >i ••'
D
V-
*,'"
AM
^
"«
'-«
. \
~<^
.»
■^,
•- *k-
\.
cs
42
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
nell, 1976; Theilacker, 1978). These techniques will suffice for
the examination of soft tissue morphology given adequately
fixed specimens. To avoid their loss, small specimens may be
prestained with borax-carmine before embedding and section-
ing; this stain can be washed out before subsequent histological
staining (Engen, 1968).
Plastic embedding (Bennett et al., 1976) is advantageous for
examination of small delicate structures, for precise records of
specimen orientation and section plane, and for the resolution
of fine cellular detail. Glycol methacrylate (Bennett etal., 1976),
epoxy resins (Humason, 1979), and other low viscosity plastics
(Hulet, 1978; L. R. White resin, London Resin Company Lim-
ited) are useful embedding media. Small specimens that can
become indistinguishable or even lost in paraffin blocks can be
easily observed in the plastic block during sectioning. As whole
mounts, specimens can be examined, measured, and meristic
characters counted before sectioning (Hulet, 1978). Techniques
developed by Ruddell (in press) reduce swelling of tissues, an
artifact sometimes encountered with glycol methacrylate
embedding. While the spectrum of histological and histochem-
ical stains applicable to plastic sections is somewhat limited,
toluidine blue counter stained with acid fuchsin has staining
reactions analogous to the more commonly used hematoxylin
and eosin. Other stain combinations also are applicable to larval
tissue embedded in glycol methacrylate (for examples see Go-
voni, 1980; Govoni et al., 1984): alkali blue 68 counter stained
with neutral red reveals fine cellular structure; VanGiesen's
picric acid counter stained with acid fuchsin reveals collagenous
fibers, the anlagen of actinotrichia; periodic acid-Schiff reagent
reacts strongly with acid mucopolysaccharides, including chon-
dromucin, and can be used to reveal cartilaginous precursors of
cartilage (endochondral) bone; alizarin red S reacts with Ca + +
ions and can reveal both calcified cartilage and bone.
Examples of Application
Histological preparations may serve the systematist in two
ways: by clarifying tissue composition and by resolving struc-
ture, thereby allowing for the determination of ontogenetic pres-
ence or absence of tissues and by offering comparisons of tissue
organization among taxa.
An example of the first use is in the identification of cartilage
and bone. The literature is replete with errors that result from
the naive interpretation of alcian blue and alizarin red S reac-
tions with cartilage and bone tissue in whole mounts. Alcian
blue reacts histochemically with the sulfate and carboxyl groups
of mucopolysaccharides (Pearse, 1968) including chondromu-
cin, the ground substance of cartilage, but it may also react with
developing bone matrices, which are rich in mucopolysaccha-
rides as well (Belanger, 1973). An alcian blue reaction, therefore,
may indicate cartilage when developing membrane (dermal) bone
is present. The reaction of alizarin red S with calcium ions
(Pearse, 1968) may indicate calcified cartilage as well as true
bone. While the clearing and staining of skeletal elements re-
mains a powerful tool (Potthoff, this volume), histological prep-
arations can clarify the identity of cartilage and bone tissue in
extremely small specimens wherein their identity may not be
clear in whole mounts.
To date, comparisons of larval fish characters revealed by
histological techniques have not been extensive and examples
of application are few. Comparative histological sections of elo-
pomorph and clupeomorph larvae illustrate the unique char-
acter of the elopomorph leptocephalus (Smith, this volume).
The unique configuration of organs and tissues is apparently
inclusive of anguilliform, elopiform, and notocanthiform lep-
tocephali. Inasmuch as Hulet (1978) also found peculiarities in
the kidney structure of the eel leptocephalus that may be unique
among vertebrates, the kidney structure of anguilliform lepto-
cephali should be compared with that of other elopomorph
leptocephali. Transient, hyaline plates occur in the basal end of
the outer integumentary epithelium of some clupeiform larvae
(Jones et al., 1966; Lasker and Threadgold, 1968; O'Connell,
1981a; Fig. 17 A), but this feature was not mentioned in the
integumentary descriptions of anguilliforms (Hulet, 1978) and
pleuronectiforms (Wellings and Brown, 1969; Roberts et al.,
1973), nor is it apparent in the perciform Leiostomus xanthunis
(Fig. 1 7B). These plates presumably function as osmotic barriers
(O'Connell, 1981a), but their systematic presence or absence is
not completely established and remains unexplained. The or-
ganization of axial musculature is another histological difference
among higher taxa. The two-layered musculature of clupeiform
larvae is aligned in opposing directions within myotomal seg-
ments (Blaxter, 1969b; O'Connell, 1981a; Fig. 17C), whereas
in perciform larvae the orientation of axial muscle fibers is
closely parallel (O'Connell, 1981a; Fig. 1 7D); this difference may
have a functional basis related to gross body form and swimming
postures (O'Connell, 1981a).
An example of the use of histological preparations to compare
microanatomical characters is the differences exhibited in elon-
gate dorsal fin rays. Elongate dorsal fin rays are features of many
unrelated taxa offish larvae (Moser, 1981), but the microana-
tomical structure of these homologous derivatives differs among
taxa (Govoni et al., 1984). A major difference is the bilateral,
paired, collagenous supporting elements of the carapid elongate
ray, as in Echiodon dawsoni (Fig. 1 7E), and the singular supports
of elongate rays of the bregmacerotid Bregmaceros atlanticus
(Fig. 17F) and the serranid Liopropoma (Kotthaus, 1970).
Monophyly in carapids has been inferred, in part, from the
distinctiveness of this synapomorphy, the elongate first dorsal
ray of their highly specialized larvae (OIney and Markle, 1979;
Markle and OIney, 1980; Gordon et al., this volume).
The often remarkable similiarity of cells and tissues, even
among phyla (Andrew, 1959), and the development of tissues
from the undifferentiated to the complex, may limit the use of
a histological approach to systematics. Yet, the unusual diversity
that characterizes ontogenetic patterns of fishes (Wourms and
Whitt, 1981), and some apparent contrasts in tissue organiza-
tion and composition that correlate with current supraordinal
classification, make histological comparisons tenable. The pre-
ceding examples of tissue and microanatomical dissimilarities
may serve to illustrate the kinds of comparisons that may prove
useful in inferring relationships as more information becomes
available. Histological techniques may provide a potentially
useful tool to the systematist; more comparative work is clearly
warranted.
National Marine Fisheries Service, Southeast Fisheries
Center, Beaufort Laboratory, Beaufort. North
Carolina 28516.
Scanning Electron Microscopy
G. W. BOEHLERT
SCANNING electron microscopy is an ideal tool for descrip-
tion of microstructure in taxonomic studies. The scanning
electron microscope (SEM) provides a surface image character-
ized by high resolution and depth of field and a three-dimen-
sional quality unavailable with other techniques. In many cases
this allows one to objectively describe microstructure where only
subjective descriptions were available in the past. It is the pur-
pose of this contribution to describe the techniques and use of
scanning electron microscopy and its application to systematic
investigations of fish eggs and larvae.
The SEM has been used in a wide variety of systematic and
evolutionary investigations. With available magnifications from
10 to greater than 100,000 times, the SEM covers the range
from dissecting and compound light microscopy to transmission
electron microscopes. It has thus been immensely important to
progress in classification in the study of micropaleontology, bot-
any, insects and mites, and a wide variety of microorganisms,
among other taxa (Heywood, 1971; Kormandy, 1975). Taxo-
nomic applications of the SEM to fishes have been more limited.
Several studies have used the SEM for studies of morphology,
including epidermis, gill tissue, optic capsules, eggs, sperm, and
embryosof fishes (Dobbs, 1974, 1975).
Microstructural analysis of otoliths of fishes with the SEM is
now common (Pannella, 1 980). For early life history stages, the
most frequent use in identification and classification has been
with the egg stage. The chorion, or external membrane, of many
species is variously ornamented with filaments, spines, patterns
of ridges, loops, blebs, and pustules ( Ahlstrom and Moser, 1 980;
Robertson, 1981; Matarese and Sandknop, this volume). These
ornamentations and the ultrastructure of the chorion are species-
specific (I vankov and Kurdyayeva, l973;Lonning, 1972). While
many of these structures may be easily visualized with light
microscopy (Hubbs and Kampa, 1946; Kovalevskaya, 1982),
the SEM often provides the best means of adequately describing
structures which are very small or transparent under the light
microscope. The egg chorion of Maurolicus muelleri, for ex-
ample, was described as "drawn up into hexagonally arranged
points," by Robertson (1976) based upon light microscopy but
as "drawn up into hexagonal ridges . . . and slightly raised at
the point of intersection" under the SEM (Robertson, 1981).
Similarly, Boyd and Simmonds ( 1 974), among others, suggested
that the chorion of southern populations of Fundulus fietero-
clitus lacked fibrils using light microscopy, whereas the SEM
showed the presence of numerous short and thin fibrils (Brum-
mett and Dumont, 1981). Thus for purposes of classification,
the SEM allows visualization of surface structures that are dif-
ficult to describe with light microscopy.
Methodology
Preparation of biological material for examination under the
SEM is concerned with preservation, dehydration, and coating
with a conductive material. Fixation of labile biological speci-
mens is necessary because removal of water during the stages
of dehydration may result in collapse of cells and other artifacts.
Depending upon the method of fixation and dehydration, the
artifacts can range from shrinkage to collapse or fracture of the
structures to be observed. It is preferable to begin with fresh,
live material. For eggs this requires either laboratory spawning
or abundant eggs from the field which can be reliably collected.
For larvae at different stages, it is diflicult without laboratory
rearing facilities. Results with formalin-fixed material from
plankton collections will generally be satisfactory for lower mag-
nification analysis of surface morphology, but may not reflect
the quality of freshly prepared material.
Fresh material should be fixed for electron microscopy. Larval
stages may first be relaxed in anesthetant solution (such as MS-
222). Initial fixatives for both eggs and larvae are generally based
upon glutaraldehyde, with concentrations ranging from 0.5 to
4.0%; lower concentrations are typically followed by post-fix-
ation. A fixative which I have found acceptable is that from
Dobbs (1974) as follows: 70% glutaraldehyde-2.0 ml, flounder
saline— 34 ml, and distilled water— 34 ml. The flounder saline
follows Forster and Hong (1958) and contains the following (in
grams per liter): NaCl, 7.890; KCl, 0.186; CaCK, 0.167; MgCK-
6H,0, 0.203; NaH,FO,H_,0, 0.069; NaHCO,, 0.84. The fix-
ative has a final osmolarity of 380 mOsm/l. Fixation should be
for 24 hours. Other authors provide several other fixatives. One
suggested by Stehr and Hawkes (1979), while more difficult to
prepare, is also useful should transmission electron microscopy
be desired for the same material. Post-fixation in osmium te-
troxide is recommended by several authors as a means of hard-
ening particularly soft tissues. Generally, 1-2% osmium tetrox-
ide in buffered saline is used. I have found this unnecessary with
fish eggs and larvae, as suggested by Dobbs (1974) and Stehr
and Hawkes (1979). It may be considered, however, if collapse
is a problem. Lonning and Hagstrom (1975) suggested that egg
chorions not post-fixed would rupture under the electron beam;
I have not noticed this.
It is the process of dehydration where the greatest artifacts
are likely to occur. With larvae, shrinkage of tissue may occur,
while eggs may suffer complete collapse. On larger eggs, punc-
turing the chorion with a sharpened dissecting needle may fa-
cilitate transfer of fluids and prevent this collapse (Stehr and
Hawkes, 1979).
Removal of water from the tissues is prerequisite to coating
and observation, which are both conducted under high vacuum.
Two methods are available, freeze drying and critical point
drying. For freeze drying, unfixed fresh material may be used.
Fixed material should first be rinsed with distilled water to
remove salts, and then plunged with little adhering water into
liquid nitrogen. Damage here may result from formation of ice
crystals if freezing rate is too slow, but this is typically not a
problem with small eggs and larvae in liquid nitrogen. Boyde
and Wood (1969) recommend using 20 ml chloroform per liter
of distilled water to increase nucleation rates and decrease ice
crystal formation. After freezing, the material is immediately
43
44
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
introduced into the freeze dryer, where water subUmes, leaving
the specimen dry and intact. Critical point drying, on the other
hand, requires dehydration through a graded series of alcohols
(20% for 24 h. then 10-20 min each in 50%, 70%, 80%, 90%,
95%, and two changes of absolute ethanol). The ethanol is then
replaced with either freon or acetone depending on whether
freon or carbon dioxide critical point dryers are used. The steps
of dehydration and transfer can be done in small specimen
holders to minimize handling and possible surface damage. Af-
ter dehydration, specimens must be mounted on SEM studs
with any of several available adhesives and tapes. The dried
specimens are particularly delicate and should be handled with
a small camel-hair brush to avoid damage to the surface. They
are then oriented onto the stud under a dissecting microscope.
Before coating, no further preparation is necessary with larvae,
but eggs have only a small area of electrical contact with the
stud. It is therefore advisable to use a conductive adhesive (such
as silver paint) to make a more complete electrical connection
and prevent charging, which decreases image quality. This paint
should be allowed to become tacky prior to positioning the eggs,
or it may cover portions of the egg itself Finally, specimens are
coated with a thin conductive layer, typically of gold or gold-
palladium, by either vacuum evaporation or ion sputtering, prior
to viewing on the SEM. At most facilities, trained SEM tech-
nicians are available; their advice and assistance are invaluable
and should be sought.
Results and Discussion
Shrinkage and other artifacts will vary depending upon the
type of material, preservation, and method of dehydration. For
fresh material preserved in a mixture of formalin, glutaralde-
hyde, and acrolein, Stehr and Hawkes ( 1979) observed a shrink-
age of approximately 10% in the eggs of Platichthys stellatus
and Oncorhynchus gorbuscha; the latter had been punctured
prior to dehydration. In the present study, eggs of Maurolicus
muellen initially preserved in 5% buffered formalin showed
varying degrees of shrinkage and collapse depending upon sub-
sequent treatment. The least shrinkage (12%, Fig. 18B) was
noted in material which was freeze dried, whereas post-fixation
and dehydration through freon 1 1 3 associated with critical point
drying resulted in shrinkage of up to 67% of the original diameter
(Fig. 18D). Eggs of this species show a hexagonal sculpturing;
under the light microscope the sculpturing is hyaline and difficult
to interpret (Fig. 18A). Eggs prepared by freeze drying clearly
show the surface sculpturing; note particularly the ridges, which
are more clearly defined (Fig. 188). For comparison, an egg
which had partially collapsed during dehydration is shown (Fig.
18D). The obvious differences in shrinkage point out the im-
portance of specifying method, initial size, and shrinkage values,
particularly for comparative or taxonomic studies.
Eggs from other species are shown to give an idea of the range
of chorion structures which may be observed. The hexagonal
pattern on M. muellen overlies a highly porous surface structure
Fig. 18. (A) Egg of Maurolicus muellen from off South Africa taken under the compound light microscope with transmitted, polarized light.
Note the emphasis of the points on the hyaUne chorion, which represent the intersections of ridges. Bar = 100 ^m. (B) Egg of A/, muellen under
the scanning electron microscope. Note the areas between what one would interpret as points on Figure 18A. which are now seen as polygonal
facets or ridges. Bar = 500 nm. (C) Individual facet of the egg of At. muellen. Note the porous and diaphanous nature of the egg surface. Bar =
50 Mm. (D) Egg of A/, muelleri post-fixed in osmium tetroxide and critical point dried. The shrinkage of this specimen is approximately 65%.
Note the differences in morphology of the ridges and surface of the egg. Bar = 100 /jm. (E) E^of Pleuronichlhys coenosus. The facets are relatively
small by comparison with M. muellen and the pattern units are more regularly hexagonal. Bar = 100 Mm. (F) Detail of two hexagons from the
egg of P. coenosus. Note the morphological differences between both the ridges and chorion surface as compared to M. muellen. Bar = 10 Mm.
Fig. 19. (A) Egg of Alherinopsis californiensis. The filaments are single, terminate in loose ends, and are distributed over the entire egg surface.
Bar = 1 ,000 Mm. (B) Egg of .-itherinops affiiUs. The egg of this species is characterized by filaments which are looped, with no free ends (Curless,
1979). This differentiates it from the egg of ,-1. californiensis, as do filament length, abundance, and basal morphology. Closed-loop filaments have
also been noted in .Aniennanus caudimaculatus eggs by Pietsch and Grobecker ( 1 980). Bar = 1 ,000 Mm. (C) Chorion of Paracaltionymus costatus
collected off South Africa. The surface features are irregular and cover the entire egg surface. This differs from species of Callionymus. which
have hexagonal patterns. Bar = 10 Mm. (D) Chorion surface of Mugil cephalus. These structures are irregular and cover the entire egg surface.
Note the superficial similarity to Paracallionymus. Bar = 10 Mm. (E) Chorion surface of an advanced ovarian egg of Coryphaenoides filifer. Note
that the surface "blebs" are arranged in hexagonal patterns and may be the precursors of a hexagonal pattern typical on eggs in this family. The
pelagic egg of this species has not been described. Bar = 10 Mm. (F) Chorion surface of an advanced ovarian egg of Coryphaenoides acrolepis.
The hexagonal ridges are better developed than in Fig. I9E. There are holes under the ndges between the intersections, which might indicate that
this species, whose egg is also undescribed, may have the hexagonal network supported on "stills" as described for eggs of Coelorhynchus spp.
(Robertson. 1981; Sanzo, 1933a). Bar = 10 Mm.
Fig. 20. (A) Spines on the chorion surface o( Oxyporhamphus microplerus. These are distributed over the entire surface of the egg. Bar = 100
Mm. (B) Chorion surface from Scomhereso.x saurus collected off South Africa. The tufts are characterized by a relatively complex basal morphology
and depending upon method of fixation, may resemble small bundles of hairs or, as here, simply coalesced tufts. Bar = 10 Mm. (C) Micropyle
and associated pores of the egg of Laclona diaphana from the Eastern Tropical Pacific. The pores shown here are restricted to this region around
the micropyle and appear to penetrate the outer layer of the chorion. Bar = 50 Mm. (D) Secondary, smaller pit structures on the remainder of the
egg of Laclona diaphana. I refer to these depressions as "pits" because closer examination does not reveal penetration through any layer of the
chorion, as opposed to the pores surrounding the micropyle in 20C. Bar = 1 Mm. (E) Head region of a larval Sebasles melanops shortly after
parturition. Polygonal epidermal cells may be noted on some parts of the body. Bar = 100 Mm. (F) Epidermis on the dorsal surface, just posterior
to the head, on an embryonic S. melanops approximately 28 days post fertilization. Note the distinct microndges and cell borders characteristic
of developing teleost epidermis. Bar = 10 Mm.
BOEHLERT: SCANNING ELECTRON MICROSCOPY
45
•/N .\!^^V-U
46
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
K
BOEHLERT: SCANNING ELECTRON MICROSCOPY
47
48
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
(Fig. 18C) as compared to that oi Pleuronichthys coenosus (Fig.
18E, F). Here, the hexagons are not only smaller, but the area
within the facets does not appear porous. SEM was used for this
species and its congeners for egg description by Sumida et al.
(1979). It is interesting to note that these authors discussed the
similarity in chorion structure of Plenronichthys spp. with that
oi Synodus lucioceps. While there were slight differences in sizes
of the polygons, the superficial similarity of chorion structure
on these phylogenetically distant genera supports a functional
role (Robertson, 1981) and independent derivation. In this in-
stance, however, SEM was valuable for understanding and in-
terpreting the differences between species and genera subse-
quently observed under the light microscope (Sumida et al.,
1979). Similarly, Keevin et al. (1980) used chorion ornamen-
tation to distinguish among genera of killifishes.
Other ornamentations include more random ridges (Para-
callionymus costatus. Fig. 19C, and Mugil cephalus. Fig. 19D),
filaments of varied length, diameter, and base morphology (Ath-
erinopsis califormensis and Athehnops affinis. Fig. 19A, B; see
also Hubbs and Kampa, 1946), tufts (Scomberesox saurus. Fig.
20B), spines (Oxyporhamphus microptents. Fig. 20A), and pits
and pores (Lactoria diaphana. Fig. 20C, D). In thecallionymids,
the small eggs of species of Callionynms have hexagonal sculp-
turing similar to that oi Pleuronichthys (Fig. 18E). In Paracal-
lionymus costatus (Fig. 19C), however, random ridges similar
to those in Mugil cephalus are apparent.
Since chorion microstruclure is formed by follicle cells during
oogenesis (Sponaugle and Wourms, 1979; Stehr, 1979), patterns
may also be discerned in ovarian eggs. The pelagic eggs of mac-
rourids are poorly known but have been described for selected
species by Sanzo ( 1 933a), Robertson ( 1 98 1 ), and Grigor'ev and
Serebryakov (1981). For Pacific species of Coryphaenoides. pe-
lagic eggs remain poorly known but apparently have hexagonal
patterns as in other members of the genus; this, is clearly shown
in ovarian eggs near the maximum size observed by Stein and
Pearcy (1982; Fig. 19E, F). Thus SEM of developing ovarian
eggs may be used to discern differences which then aid in iden-
tification of eggs from plankton samples.
For larval stages, SEM has been used for the description of
development of several surface structures, such as the olfactory
organ (Elston et al., 1981) and lateral line neuromasts (Dobbs,
1974). For taxonomic studies, differentiation of fine-scale mor-
phological differences, such as dentition or fine-scale spine ser-
ration, may be useful. Its most valuable use may therefore be
for later larval development, since pigmentation and other char-
acteristics in early larvae are better seen with conventional
methods (Fig. 20E, F).
To conclude, SEM may serve as an adj uct to traditional meth-
ods in the description of fine structure in fish eggs and larvae.
For high magnification, high resolution visualization of surface
morphology, it remains the most effective tool available. Under
lower magnifications, it may allow one to clearly visualize struc-
tures which are difficult to interpret using standard microscop-
ical methods (Fig. 1 8A, B).
Oregon State University , Marine Science Center, Newport,
Oregon 97365.
Developmental Osteology
J. R. Dunn
ONE legacy left by Elbert H. Ahlstrom was an appreciation
of the value of developmental osteology of teleosts as a
taxonomic aid and as an indicator of phylogenetic affinities.
Although numerous studies have been made on the growth of
various bones in teleosts, such descriptions have not been widely
used in assessing relationships of fishes. I have recently re-
viewed, in some depth, the application of developmental os-
teology in taxonomic and systematic studies of teleosl larvae
(Dunn, 1983b). Here I present a brief overview of some skeletal
structures in teleosts whose ontogeny offers potential utility in
inferring phylogenetic affinities. It is hoped that this precis will
encourage ichthyologists to examine the development of bones
in the course of their systematic studies.
Ontogenetic Changes in Skeletal
Structures
Cranial and associated bones— CTaniaX osteology has, of course,
been the foundation of systematic studies of adult fishes, but
the development of cranial bones has been little used in phy-
logenetic studies. Numerous descriptions of the ontogeny of
cranial bones exist in the literature (e.g., Bhargava, 1958; Bert-
mar, 1959; Kadam, 1961; Weisel, 1967; Moser and Ahlstrom,
l970;Mook, l977;Leiby, 1979b; Yuschak, 1982). Additionally,
the sequence of ossification of head bones has been described
for a variety of taxa (e.g., Moser, 1972; Aprieto, 1974; Leiby,
1979a; Dunn, 1983a; Kendall and Vinter, 1984). The devel-
opment of certain cranial structures has also been shown to be
of taxonomic value (Fritzsche and Johnson, 1980), yet com-
parative studies of the developmental osteology of the skull of
related groups of teleosts seem rare (e.g., Norman. 1926b; De
Beer, 1937).
Available evidence suggests that the sequence of ossification
of the skull of teleosts is a conservative (i.e., relatively constant
among different phyletic groups) process (De Beer, 1 937; Mook,
1977). Among the bones which ossify first are those in areas of
high stress, such as feeding (jaw bones) and respiration (bran-
chial region), as noted by De Beer ( 1 937), Weisel ( 1 967), Moser
and Ahlstrom (1970), Mook (1977), Yuschak (1982).
Examples of ontogenetic changes in skull bones which suggest
that these structures might offer insight into phylogenetic affin-
DUNN: DEVELOPMENTAL OSTEOLOGY
49
ities include upper jaw bones (Berry, 1 964a), head spines (Ken-
dall, 1979; Washington, 1981; Yuschak, 1982; Washington and
Richardson, MS), gill arches (Leiby, 1979b; Yuschak, 1982;
PotthofTet al., 1984), and lateral skull bones (Leiby, 1979b).
Patterns of chondrification may also be of value in inferring
phylogenetic relationships. Washington and Richardson (MS)
noted that while chondrification of skeletal bones in most scor-
paeniform fishes is a relatively brief process, occurring in pre-
flexion and early flexion larvae, chondrification was prolonged
(occurring through most larval development) in hexagrammids
and in three genera of cottids. These authors also considered a
unique pattern of ossification of cartilaginous rings in the regions
of the parietal and frontal spines as a synapomorphic character
uniting three genera of cottids.
Vertebral column and associated bones. — Vertebral centra, neural
and haemal spines, apophyses, and ribs all undergo variable
changes in configuration with growth. A number of workers have
documented the development of the vertebral column and as-
sociated bones in a variety of taxa, but attempts have not been
made to analyze the phylogenetic significance of the ontogeny
of these structures. The sequence and direction of ossification
of vertebral centra is known to vary among taxa (e.g., Moser
and Ahlstrom, 1970; Mook, 1977; Potthoff" et al., 1984), but
this character has yet to be analyzed among groups of fishes.
Among those elements of the vertebral column which have
been studied in various taxa, Potthoff"and Kelley (1982) noted
that the neural and haemal arches in Xiphias first develop dis-
tally opened, whereas in other perciforms studied, split arches
were observed in small larvae on the anterior two centra only.
Washington and Richardson (MS), in their study of cottid larvae
and scorpaeniform outgroups, noted in various taxa the reduc-
tion or absence of the first neural spine, presence or absence of
autogenous neural arches on centrum one, shape of anterior
neural arches, and whether or not the first neural arch was
distally fused or open. Potthoff" and Kelley (1982) cited the
unique position and development of ribs in Xiphias compared
to other perciforms studied, and Washington and Richardson
(MS) examined the location, number, and position of ribs in
cottids and perciform outgroups.
Fins and their supports— Y>OTsaX and anal fins— The sequence
of formation of dorsal and anal fins as well as the order of
development of their constituent spines and/or rays varies among
taxa (Dunn, 1983b). This succession of formation may be rel-
atively constant among related groups or it may vary, but the
phylogenetic significance of these events, if any, has yet to be
analyzed. Additionally, numerous taxa of larvae possess tran-
sient, often bizzare, structures, such as elongate dorsal spines
or rays or anal rays (e.g., Kendall, 1979; Moser, 1981). These
structures are of taxonomic value and may contain phylogenetic
information, but the homologies of these structures, if any, are
not known (Govoni, this volume).
PotthoflTet al. (1984) indicated that the second dorsal and
anal fins are the first to develop in most perciform fishes. How-
ever, in generally more advanced species, dorsal fin rays (or
spines) develop first anteriorly and second dorsal and anal fin
ray development starts after the first dorsal fin is either partially
or fully developed. Fahay and Markle (this volume) described
the sequence of fin formation in gadiform fishes. Usually the
vertical fins ossify at nearly the same time, but two or more
centers of ossification are present in those genera (e.g., Molva.
Merluccius) with a single long dorsal fin (or a short first dorsal
fin preceding a longer second dorsal fin).
The ontogeny of pterygiophores has received considerable
attention from Potthofl"and colleagues (e.g., PotthofT. 1975, 1980;
Potthoff'et al., 1980, 1984). The developmental pattern of fin
pterygiophores may suggest phylogenetic relationships. PotthofT
and Kelley (1982) noted that the first dorsal pterygiophore in
Xiphias arose from either one or two pieces of cartilage, as is
the case in Morone (Fritzsche and Johnson, 1 980), but not in
scombrids. Washington and Richardson (MS) observed the on-
togenetic migration of dorsal fin pterygiophores, relative to neu-
ral arch position, in three cottid genera. Proximal and distal
radials may fuse during ontogeny (Yuschak, 1982) and the pres-
ence or absence of medial radials may characterize certain groups
of fishes (PotthofT and Kelley, 1982).
Pectoral and pelvic fins and their supports.— 'Wilh some excep-
tions, pectoral fins develop rays later in the larval period than
median fins (Dunn, 1983b). Transient, elongate spines and
rays also develop in the pectoral fins of some taxa (Moser and
Ahlstrom, 1974; Moser, 1981); such structures may be of taxo-
nomic value, but their phylogenetic significance, if any, and their
homologies are not known. Relatively few descriptions have
been published on the development of the pectoral fin (e.g.,
Houdeand PotthofT, 1976; Potthoff", 1980; Potthoff"and Kelley,
1982; Yuschak, 1982; Potthofl["et al., 1984), and few systematic
inferences have been drawn. PotthofTet al. (1984) noted, in
Anisotremus virginicus. the ontogenetic fusion of the supratem-
poral-intertemporal, the elongation of the anterior coraco-scap-
ular cartilage, and the reduction in length of the posterior pro-
cess. Washington and Richardson (MS) examined the orientation
of the cleithrum, as well as its outer lip, the length of the scapula-
coracoid complex, the base of the cleithrum, and the cleithral
extension over the pelvic bone (among other characters of the
pectoral girdle) in their analyses of cottids and their allies.
The ontogeny of the pelvic fin and its supporting structures
also has been little investigated (PotthofT, 1980; PotthoflTet al.,
1980; Fritzsche and Johnson, 1980) and infrequently used in
systematic studies. Dunn and Matarese (this volume) indicated
that in gadid larvae the length of the posterior-lateral process
of the basipterygia differed among subfamilies and tended to be
reduced or wanting in those genera presently considered ad-
vanced.
Caudal fin.— The development of the caudal fin in teleosts, a
subject Dr. Ahlstrom was extremely interested in (e.g., Ahlstrom
and Moser, 1976), seems to have received more study than other
bony structures. However, few workers have attempted to in-
terpret the phylogenetic significance of the development of this
fin (Dunn, 1983b).
The fusion of bones, reduction in size of structures, or'loss
of elements by absorption can frequently be observed in the
development of the caudal fin in some fishes. Additionally, based
on ontogenetic evidence, the structure of this fin may differ from
that commonly accepted based on adult specimens (Dunn,
1983b).
Ontogenetic changes in the caudal fin and associated bones
which have been used to infer phylogenetic relationships include
the reduction through fusion of ural centra (Moser and Ahl-
strom, 1 970; and others), discreet or fused hypural bones (Wash-
ington and Richardson, MS; Dunn and Matarese, this volume),
absence of the parhypural in certain taxa which normally possess
one (Washington and Richardson, MS), characteristics (e.g..
50
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
shape, modification, autogenous or fused to the centra) of neural
and haemal spines on preural centra associated with the caudal
fin (Washington and Richardson, MS; Dunn and Matarese, this
volume), and number of vertebral centra supporting the caudal
fin (Washington and Richardson, MS; Fahay and Markle, this
volume).
Attention has recently been directed toward the presence of
radial cartilages (their position and shape during development)
in the caudal fin of certain teleosts (Kendall'; PotthofT et al.,
1984). These structures may contain information of value in
assessing phylogenetic relationships.
Squamation.—The development of scales in teleosts has been
described for a variety of taxa (e.g.. Berry, 1960; Burdak, 1969;
Fujita, 1971; White, 1977; Potthofl'and Kelley, 1982). The se-
quence of development of scales and their origin on the fish
differs among taxa, and scales undergo changes with ontogeny
(e.g.. White, 1977; Potthoffand Kelley, 1982). The acquisition
' Kendall, A. W., Jr. 1981. Ventral caudal radials— oft overlooked
structures. (Paper presented at annual meeting Amer. Soc. Ichthyol.
Herpetol., Corvallis, OR, June 1981; Abstract in Copeia 1981:935).
of scales on fish usually occurs during their transformation to
the juvenile stage; however, a number of groups (e.g., Zaniolepis.
serranids, holocentrids, and xiphiids) acquire scales during the
larval period. Such developmental changes have apparently not
been analyzed among diverse groups of fishes.
Perspective
Developmental osteology of teleosts appears to be an under-
exploited approach of potential value in increasing our under-
standing of the relationships of fishes. Studies of developmental
osteology of teleosts may contribute much to our understanding
of homology, the central concept of all biological comparisons
(Inglis, 1966; Bock, 1969; Wake, 1979) in our search for prim-
itive and derived character states. A number of investigators
present at this symposium are actively engaged in evaluating
ontogenetic changes in ossified structures in their studies of
various taxa of larval fishes. An appraisal of this method may
well be in the future, but evidence provided during the course
of this meeting will contribute to such an evaluation.
Northwest and Alaska Fisheries Center, National Marine
Fisheries Service, 2725 Montlake Boulevard East,
Seattle, Washington 981 12.
Otolith Studies
E. B. Brothers
ALTHOUGH the value of otolith studies in systematic ich-
thyology is well established, essentially all studies to date
deal with the otoliths of adults, or only incidentally juveniles,
and are usually limited to the external morphology of the typ-
ically largest otolith, the sagitta (see reviews of Weiler, 1968;
Casteel, 1974; Hecht, 1978; Huygebaert and Nolf, 1979). Oto-
liths of larvae, which are of recent interest in terms of age,
growth, mortality, and life history studies (Brothers et al., 1976;
Struhsaker and Uchiyama, 1976; Methot and Kramer, 1979;
Townsend and Graham, 1981; Kendall and Gordon, 1981; La-
roche et al., 1982; Lough et al., 1982; Bailey, 1982; Brothers et
al., 1983) have been ignored from a taxonomic point of view.
This is perhaps not surprising due to their very small size and
generally simpler form, with an apparent lack of obvious dis-
tinguishing external features. Although the internal structure of
larval otoliths appears to be more variable than the external
form, no comparative taxonomic studies have been attempted
to date. In addition, relatively little has been done on compar-
isons of these features of adult otoliths, noting that in a real
sense, the internal anatomy of the adult otolith is just the cu-
mulative historical record of ontogenetic changes in external
structure and growth patterns. Comparative studies on features
other than external appearance have tended to be at the crys-
tallographic, mineraiogical and chemical level. Carlstrom's ( 1 963)
research on the crystallographic structure of fish otoliths and
otoconia was a pioneering attempt to apply structural and com-
positional information to understanding the broad outlines of
vertebrate evolution. A few studies have followed this line of
investigation (Lowenstam, 1980, 1981; Lowenstam and Fitch,
1978, 1981), however the discrimination ability of crystallo-
graphic techniques is certain to be limited by the relatively few
crystalline varieties known to exist in ear stones. Analysis of
the amino acid composition of the major organic fraction of
otoliths (Degens et al., 1969) offers another possibility for taxo-
nomic information, however it is unlikely to be useful for spe-
cific identification of individuals. Finally, trace element analysis
of otoliths (Gauldie et al., 1980; Papadopoulou et al., 1978,
1980) may allow for stock and perhaps species discrimination,
but again the small sample sizes offered by larval otoliths impose
severe or impossible methodological problems unless x-ray mi-
croprobes or ion microscopes are employed. New analytic tools
for chemical studies could offer unique insights into fish sys-
tematics.
Recently renewed interest in fish otoliths, due primarily to
the recognition of daily growth increments (Pannella, 1971,
1980). has resulted in an expanding effort toward collecting,
examining and cataloging the otoliths of larval fishes. As we
begin to study the external and internal structure of this material
for systematically useful characters, we should begin to develop
a new set of morphological criteria for species identification,
taxonomic relationships, and perhaps phylogenetic reconstruc-
tion.
BROTHERS: OTOLITH STUDIES
51
Fig. 2 1 Abrupt changes in external and internal morphology of the sagitta associated with the end of the larval stage. (A) Scanning electron
micrograph of the medial face of the left sagitta (9 mm SL) of a french grunt {Haemuton flavohtwatum). (B) 12 mm SL, showing development
of "secondary growth centers." (C) Enlargement of area in previous specimen. (D) 44 mm SL. Scale omitted: 12 mm = 500 ixm. (E) SEM of
ground and etched hake (Merluccius sp.) sagitta. showing growth centers around the larval otolith. (F) Photomicrograph of ground sagitta of a
largemouth bass, Micropterus salmoides. The larval portion of the otolith is in the lower right comer.
52
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 22. Photomicrographs of otoconia in teleosts. (A) Bonefish, Albula vulpes. free otoconia. (B) Bonefish, otoconia embedded m the sagitta.
General Methodology
The otoliths (sagittae, lapilli, and asterisci) of larval fish are
usually the first calcified structures to appear in the development
of an individual. At least some of the otoliths are frequently
evident before hatching. Over the larval life, they vary in size
from a few to several hundred micra for different taxa and ages.
Because of their composition and small size (high surface to
volume ratio), larval otoliths are very sensitive to degradation,
decalcification, and dissolution in acidic solutions (McMahon
and Tash, 1979), and great care must be exercised in preserving
larval fish and otoliths. Improper handling results in rapid and
irreversible damage. Fresh larvae are best stored for later otolith
extraction in three ways: 1 ) frozen, 2) fixed and maintained in
strong ethanol solutions (preferably 95%), 3) dried (e.g., on glass
slides). The last technique is least preferred due to increased
difficulties in otolith removal and general damage to the larvae.
Removal from embryos and larvae involves microscopic dis-
section with fine needles. The use of crossed polarized filters is
sometimes helpful in locating the otoliths, although they are
generally clearly visible in the otocysts or otic capsules with
standard transmitted illumination. Dissection is best carried out
in water, and opaque larva can be cleared by brief exposure to
a weak KOH (1%) solution. Air dried otoliths should be trans-
ferred on the tips of oil wetted (immersion) needles, and for
light microscopy may be stored in oil on slides or permanently
mounted under coverslips with a neutral medium (non-acidic).
In the latter case, care must be taken to prevent the otoliths
from being cracked or crushed as the mounting medium shrinks
and pulls down the coverslip. In most cases larval otoliths are
small and thin enough to preclude a need for grinding. Light
microscopy is best applied to studies of internal structures, al-
though some external features can be viewed with either surface
microscopy or transmitted light and wide openings of the con-
denser diaphragm. Compound microscopes should have high
quality oil immersion optics (preferably to at least 1 ,000 x ) and
polarizing filters. For the latter, a single, rotatable field polarizer
helps in resolving internal structures, while an analyzing polar-
izer can be employed to locate the very small, but highly bire-
fringent otoliths on slides. A moderately high resolution (at least
500 lines) black and white video system is an additional, but
invaluable accessory. Such a system reduces eye fatigue, sim-
plifies group viewing, measurement and photography, and most
importantly can substantially enhance image quality by elec-
tronic adjustment. It is also a necessary component in a variety
of automatic and semi-automatic image analysis systems.
Scanning electron microscopy is most useful for high reso-
lution views of external structures, for examination of fine (< 1
fim) internal features, and for confirmation of suspected optical
artifacts. However the technique is also more expensive and
time consuming and may necessitate critical preparation. Whole,
cleaned and air-dried otoliths can be mounted and coated by
standard techniques. Internal views require embedding, grind-
ing, polishing and etching before stub mounting and coating.
The most recent important development in SEM preparation
is the use of etching solutions other than the initially preferred
HCl. Haake et al. (in press) summarize a technique for SEM
preparation of larval otoliths.
Otolith Morphology and Early Ontogeny
There are a number of papers which deal with the general
structure and composition (Hickling, 1931; Degens et al., 1969:
Blackler, 1974: Pannella, 1980), mechanism of growth (Irie,
1960: Dunkelburgeretal., 1980; Campana, 1983), and functions
of the otoliths and otolith organs (Popper and Coombs, 1 980a, b;
Piatt and Popper, 1981). This work has not specifically dealt
with larvae, however the gross morphology and processes should
be comparable with older fishes.
The otic capsule or otocyst forms very early in the ontogeny
of fishes and is an obvious landmark in the head of newly
hatched larvae. The earliest evidence of the otoliths is one to
several small (usually less than 10 ixm) optically dense bodies,
referred to here as primordia. From their physical appearance
and etching properties, the primordia are assumed to be sub-
stantially composed of organic matrix (probably the fibroprotein
otolin), and are soon calcified and surrounded by an accreted
layer of calcium carbonate and matrix. There are distinct dif-
ferences between certain taxa, usually at the supraspecific level,
with regards to the morphology of the primordia. Distinctions
also exist between the sagitta, lapillus, and asteriscus, so com-
parative studies must be careful to properly identify the otoliths
examined. Variation in primordial form involves the size, shape,
and number per otolith. Surrounding the primordium (partic-
BROTHERS: OTOLITH STUDIES
53
Fig. 23. Otolith primordia and cores. (A) SEM of single primordium and core in a french grunt (Haemulon flavolineatum) lapillus. (B)
Photomicrograph of single primordium and core in a mimic blenny {Labrisomus guppyi) sagitta. (C) Multiple primordia in the lapillus of a white
sucker {Caloslomus commersoni). (D) Multiple primordia in the sagitta of a seahorse (Hippocampus sp.). (E) Multiple primordia and cores in the
lapillus of a banded killifish (Fiindulus diaphamis). (F) SEM of multiple primordia and cores in the sagitta of a rainbow trout {Salmo gairdneri).
ularly in the sagitta and lapillus) is a discrete, relatively ho-
mogeneous zone of calcified material usually delimited by a
distinct, thin, optically dense (matrix-rich) layer. This layer de-
fines the boundary of the core. In some cases, careful exami-
nation of the core may reveal diffuse, very faint, or extremely
fine growth increments, however, they are easily distinguished
from the more distinct incremental growth pattern distal to the
core. Taxonomically related differences in core size, shape and
number generally parallel differences in the primordia.
The external morphology of larval fish otoliths is much less
54
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 6. Occurrence of Multiple Primordia in Fish Otoliths (see
Text for Explanation).
Order Mormyriformes
Mormyridae
Order Salmoniformes
Esocidae
Umbridae
Salmonidae (including Coregoninae)
Osmeridae
Order Cypriniformes
Characidae
Cyprinidae
Catostomidae
Order Siluriformes
Ictaluridae
Bagridae
Order Atheriniformes
Exocoetidae
Oryziatidae
Cyprinodontidae
Belonidae
Anablepidae
Poeciliidae
Atherinidae
Order Syngnathiformes
Gasterosteidae
Syngnathidae
Order Scorpaeniformes
Cyclopteridae (Cyclopterinae and Liparinae)
Order Gobiesociformes
Gobiesocidae
Order Perci formes
Istiophoridae
Stichaeidae
Percichthyidae
Order Pleuronectiformes
Pleuronectidae
variable than seen for adults. Similarity between taxa is greatest
in the youngest and smallest individuals, in which the otoliths,
particularly the sagittae and lapilli, tend to resemble flattened
spheroids or hemispheres. Landmark features used in char-
acterizing adult otoliths such as the form of the sulcus, rostral
projections, cristae, colliculi, ostia etc. are initially not evident
or weakly developed in most fishes. Exceptions to this gener-
alization may prove to be useful taxonomic characters (e.g., in
various istiophorids, the sulcus acousticus is clearly developed
in larvae only 6 mm SL). Exaggerated or distinctive morpho-
logical features of adult otoliths of some taxa may also begin to
develop in the early larval stages. For example, if a species has
a markedly elongate sagitta, such as found in some callionymids
or fistulariids, then the larval otolith may show a tendency for
greater growth along the anterio-posterior axis. Unfortunately,
such early evidence for adult otolith characters is often not
present, particularly for the many species which show an abrupt
change in otolith growth patterns at the end of the larval phase.
Nevertheless, there are other unique or distinctive larval otolith
features in many taxa, and they are potentially valuable for
systematic studies.
Aside from shape, there are at least two other "external"
otolith characters which may be used for taxonomic work; these
involve the relative sizes and times of formation of the different
otoliths; the sagittae, lapilli, and asterisci. In certain taxa, such
as the Ostariophysi, the sagitta is highly modified from the
typical teleost condition, being smaller and very elongate; and
the asteriscus is relatively enlarged. In clupeids, the lapillus is
unusually small and distinctively shaped. Differences of this sort
exist to a lesser degree at lower taxonomic levels and may be
used in larvae for distinguishing groups. The time of appearance
of the otoliths in development is also a variable feature offish
ontogeny. Many or perhaps most species have sagittae and lapilli
at hatching, the former usually noticeably larger even at this stage.
There is a general positive relationship between egg size, time
to hatching and state of otolith development at hatching. Fishes
with very large eggs and corresponding hatching size may also
have the asterisci present at this early stage, however for the
majority of fishes, these otoliths appear later, and are sometimes
not apparent until the end of the larval stage. The asterisci are
distinctive in other respects as well; all species I have looked at
have a poorly defined core with multiple primordia; the calcium
carbonate is deposited as vaterite (Lowenstam and Fitch, 1981)
rather than the aragonite of the sagittae and lapilli; and there
are qualitative differences in the appearance of growth incre-
ments.
Internal structures other than the primordium and core may
also have direct or indirect systematic applications. It is well
documented that otoliths grow by the addition of layers which
are deposited on a diel cycle (see earlier references on larvae,
plus review by Pannella, 1 980; also Barkman, 1 978; Wilson and
Larkin, 1980; Steffensen, 1980; Victor, 1982; Victor and Broth-
ers, 1982). These daily growth increments are usually simple
bipartite structures composed of one protein-rich and one pro-
tein-poor calcareous layer. In certain situations (especially fast
growth and large otoliths) subdaily increments (formed over
shorter time intervals) of similar structure may also be present.
The timing of the production of the defining boundary of the
core, which also corresponds to the onset of incremental growth
around the core, is another "internal" character that varies be-
tween taxa. Some groups start incremental growth before hatch-
ing, others at hatching, and still others at about the time of yolk
absorption and the onset of exogenous feeding (Brothers et al.,
1976; Radtke and Waiwood, 1980; Radlke and Dean, 1982;
Radtke, 1984). There appear to be clear taxonomic trends in
these characters which are also related to other trends in egg
size and developmental rate and pattern.
Some Examples of Taxonomically Related
Trends in Larval Otolith Form:
External Morphology
The development of the general form of the adult sagitta is a
gradual process in many species, whereas in others there may
be one or more relatively abrupt changes in growth form, par-
ticularly around the time of transformation from larva to ju-
venile. This change involves the development of "secondary
growth centers" which first appear externally as angular to
rounded protuberances on the sagitta surface (Fig. 21; internal
structure is discussed below). The result of the expanding growth
around these centers is the eventual surrounding of a discrete
larval otolith and the stronger development of form and surface
characters of the adult sagitta. In examining the otoliths of over
BROTHERS: OTOLITH STUDIES
55
IOh™
10h">
B
Fig. 24. Pnmordia and cores of goby otoliths. (A) Sagilta of adult sirajo goby {Sicydiuni plumieri). (B) Sagitta from an unidentified goby larva.
100 families of fishes, this soil of sagittal growth pattern appears
to be characteristic in a number of higher level taxa (e.g., many,
but not all, perciform families; some myctophids; certain but
not all anguilloid families, pleuronectiform, gadiform and scor-
paeniform fishes; Percopsis, and others). It is not certain whether
the presence of this character is consistent enough to be used
as a diagnostic feature, and it also occurs too late in development
to be of use in larval identification. Lapilli and asterisci tend to
show more gradual changes in shape and growth (Brothers and
McFarland, 1981) and I have not observed the discontinuous
pattern described above. Lapilli undergo transitions in incre-
mental patterns at about the same time that the sagitta changes
in growth form (Brothers and McFarland, 1981; Brothers, un-
published), however these are not obviously evidenced in ex-
ternal morphology of the former.
An unusual and surprising character has been found in a
preliminary survey of several of the "lower" teleosts. This fea-
ture, the presence of otoconia in the sacculus and/or utriculus
in addition to the otoliths, has only been noted for non-teleos-
tean bony fishes, i.e., holosteans, chondrosteans, brachiopte-
rygians, dipnoans (Carlstrom, 1963) and probably Latimena
(Brothers, unpublished). Osteichthyan otoconia or statoconia
are numerous (hundreds to thousands), small (from a few to
1 00 ^m) calcareous bodies (vateritic, sometimes aragonitic) which
are found in close association with the otolith (Fig. 22). They
generally have a very characteristic lens shape, although some
may tend towards an hexagonal outline. Internal features are
variously developed; a primordium-like body is usually present
and incremental growth is seen in some. Unexpectedly, otoconia
were found in representatives of the following teleost families:
Albulidae, Congridae, Anguillidae, Muraenidae. Moringuidae,
Notopteridae, Osteoglossidae and Pantodontidae. The character
appears to be an example of a synplesiomorphy shared between
non-teleostean osteichthyans and two teleostean superorders,
and Osteoglossomorpha and the Elopomorpha. Not all species
and possibly families in the latter two groups show the character,
so apparently it has been lost independently more than once.
The presence of otoconia is usually not apparent until the early
juvenile stage, they are not seen in the few larvae I've had
available, however, their taxonomic interest warrants mention
here.
Internal Morphology
There are a number of taxonomically related trends in the
size and shape of the primordium and core of sagittae and lapilli.
Table 6 lists all the families (of 1 13 sampled) found to have
representatives with multiple or clustered primordia (inclusion
in the table does not necessarily indicate that all family members
have the character). In some, particularly the salmonids and
related families, the primordia are clearly separated and may
each be surrounded by discrete multiple cores, whereas in others,
such as the Atheriniformes and Gasterosteiformes, the multiple
primordia are more lightly grouped and are usually surrounded
by a single core (Fig. 23).
Two other primordium and core characters have been found
to be unique to certain taxa. In the gobies and related families
( 1 5 genera; Gobiidae, Microdesmidae, Eleotridae, and Gobioid-
idae) all species invariably have an elongate primordium in the
sagittae and lapilli (usually with a slight central constnction. Fig.
24) which has not been seen in any other group. Since this feature
is present at hatching, it allows for rapid and certain identifi-
cation of these speciose families. The parrotfishes (Scaridae, 4
genera examined) appear to have a family-specific early growth
pattern in the sagitta which also allows for the identification of
very young larvae. The nearly spherical primordium and core
grow asymmetrically for about the first 5 days, adding new
increments in a restricted area on the distal face before the
growth pattern changes to one producing a hemispherical larval
otolith. The result of this pattern (Fig. 25) is that the core is
clearly on a different focal plane from a section normal to the
majority of larval growth increments. The core is therefore
asymmetrically placed nearer to the proximal or internal face
of the sagitta. This feature is easily observed in whole larval
otoliths and has not been found in related families such as the
labrids, although these families share other larval otolith char-
acters.
A second class of internal features has obvious external man-
ifestations described above, although they may be distinguished
externally for only a discrete period in development. "Secondary
growth centers" appear in optical sections or SEM views as foci
for increment formation removed from the core (Fig. 2 1 ). Sp)ecies
in which otoconia occur are also found to have these bodies
56
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
IOh""
lOt""
Fig. 25. Primordia and cores of parrotfish sagittae. (A) Unidentified scarid larva, medial face up, core in focus. The dark crescent is a portion
of the crista on the surface. (B) Same as previous, but with increments in focus. (C) Suspected scarid larva, core in focus. (D) Same as (C),
increments in focus.
incorporated into the otoliths. The mechanism appears to be
that the otoconia adhering to the otolith surface are surrounded
by new material accreting on the otolith, and eventually these
"included" otoconia are found deep within the otoliths of larger
fish. In some species, such as Anguilla rostrata otoconia are
found in dense bands corresponding to annual zones. "Includ-
ed" otoconia have only been observed in juveniles or older
individuals.
Transitions in otolith microstructure involving changes in the
width and optical density of growth increments (Fig. 26) may
be related to a variety of morphological and eco-behavioral
changes in the early life history offish (Pannella, 1 980; Brothers,
1981; Brothers and McFarland, 1981; and numerous other pa-
pers; also related works by Postuma, 1974, and McKem et al.,
1974). Hatching, yolk absorption, changes in feeding and hab-
itat, postlarval transformation, and settlement can all poten-
tially influence the deposition pattern of daily and subdaily growth
increments. To the extent that life history patterns consistently
diflfer between taxa, we may expect to find microstructural evi-
dence of events in the early life history which are of systematic
value. Difierences between taxa will then be expressed as dif-
ferences in the timing of marks (e.g., hatching) and otolith tran-
sitions and in their intensity and duration. Thus we may use
otoliths to record ecological information which may then be
applied to systematic studies. An even simpler approach might
just be a quantitative comparison of growth rates as determined
from daily growth records (once validated, and the fish growth-
otolith growth relationships are known), however care should
be taken to avoid problems due to high intraspecific variability
in growth rate (e.g., Methot, 1981; Bailey, 1982; Brothers et al.,
1984). Another possibility is the use oflarval life duration as a
taxonoinic character. There is evidence to both species speci-
ficity and very limited variability in some taxa, as well as vari-
ability or flexibility in others (Brothers et al., 1983; Thresher
and Brothers, in press; Brothers and Thresher. MS.; Brothers
and Erdman, unpublished), so caution must be exercised in
using this character as a taxonomic tool.
A final ecologically related application is the determination
of spawning time (and perhaps place, by correction for current
drift) by age determination of larvae, with correction for the lag
between fertilization and increment initiation (Townsend and
Graham. 1981; McFarland et al., unpublished). When difl^er-
ences in spawning times are suspected or known to exist for
taxa, then larval age may be used to help in assigning identifi-
BROTHERS: OTOLITH STUDIES
57
IOh"
B
Fig. 26. Transitions in otolith microstructure associated with settlement and transformation from the larval to juvenile stage. (A) Striped
parrotfish (Scarus iserti) sagitta. (B) Queen angelfish (Holacanthus ciliaris) sagitta.
cation. Under the best of circumstances, when spawning is rel-
atively discrete in time, differences of only a few days could
potentially be resolved.
The last area in which otolith studies might be of value in
systematic studies is in the presentation of descriptive papers
on fish development. Until now all illustrations and descriptions
of development of wild caught larvae were related to body size
since we had no information on the age of these specimens. We
suspect, and in some cases have direct knowledge (cited earlier)
that growth rates of larvae are moderately to highly variable,
yet we have no data on the relationship between age and growth
rate and the appearance and form of standard characters such
as pigment, ossification, meristics, and morphometries. Perhaps
some of the variability seen in size specific descriptive accounts
is the result of the effects of different growth rates on the char-
acters. I urge that we should make an extra effort to determine
the age of wild-caught larvae, used in descriptive studies so we
may be able to establish age and/or growth rate specific accounts
as well as size specific ones. Of course another problem with
size is the highly variable shrinkage rates caused by handling
and preservation. Alternately we should perform laboratory ex-
periments to examine the relationship between growth rate and
developmental rate. In this way we may be able to understand
some of the underlying causes for intraspecific variation in larval
fish characters.
Section of EcoLOCiv and Systematics, Cornell University,
Ithaca, New York 14853. Present Address: 3 Sunset
West, Ithaca, New York 14850.
Preservation and Curation
R. J. Lavenberg, G. E. McGowen and R. E. Woodsum
THOSE processes by which we fix or kill living tissues without
significantly altering their gross anatomy, and preserve or
maintain these tissues on a long-term basis have routinely re-
quired the use of formalin solutions (Fink et al., MS; Markle,
1984). This certainly is the case for fish eggs and larvae. The
protocols for use of formalin as a fixative and preservative for
ichthyoplankton have been reviewed and standardized in sev-
eral techniques manuals (Ahlstrom, 1976; Castle, 1976; Smith
and Richardson, 1977). These protocols are well established and
it is not our intention to repeat them here. Rather we wish to
elaborate on some of the problems associated with preservation
and curation, and to propose recommendations to resolve those
areas of real or potential conflict.
There are two areas of special concern to us that dictated how
our investigations proceeded. First, we wish to ensure that em-
bryonic pigment is retained in both the egg and larval stages in
both the fixation and long-term preservation procedures. Sec-
ond, for ontogenetic stages of larvae we were guided by a concern
for protection of mineralized structures, guarding particularly
against their loss.
Specimens that are well-fixed and properly preserved are im-
portant not only to ichthyoplanktologists but to a broad spec-
trum of biologists, fish systematists, and museum curators.
Among fixatives, bufters and preservatives there is no unani-
mous agreement on the most appropriate ones. The problems
that plague our understanding of the processes associated with
58
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 27. A proposed method to archive the early life history stages of fishes. In the left foreground is a series of three vials, the first contains
the specimens and preservation fluid and is capped with a polyseal closure. This first vial is placed into the second with the documentation. The
third vial is a complete unit. As evaporation occurs the outer vial pops free of its plastic closure, indicating that the vial requires curatorial
attention. The vials can be placed together in commercially available paper trays, which can be arranged in commercially available wooden trays
much like entomological collections are maintained.
these chemicals and prevent us from standardizing a protocol
are not biological ones but rather those of chemistry.
Fixatives. — Formahn generally is accepted as the most appro-
priate fixative. However, it must be used in a specific concen-
tration, polymerizes with age and with contact with metals, and
is a poison. Tucker and Chester (in press) found that formalin
used with salt water causes significant shrinkage, whereas an
unbuffered 4% solution of formalin mixed with freshwater caused
the least amount of shrinkage and distortion during fixation.
They found that pigment preserves best in a solution of un-
buffered freshwater formalin. Although the pigment holds up
well in this solution, the skeleton decalcifies and reduces or may
even prevent staining for either bone or cartilage using the meth-
ods of Dingerkus and Uhler (1977). In the absence of a suitable,
inexpensive substitute we recommend that formalin be used for
fixing zooplankton samples, using the standard ichthyoplankton
protocols described by Smith and Richardson (1977). This pro-
tocol could be modified so as to use freshwater rather than
seawater in preserving the sample (Smith and Richardson, 1977:
16-section 2.1.3.1) so as to reduce shrinkage.
Buffers— The problems associated with buffers are more diffi-
cult to unravel. Buffers have been used in an attempt to control
fluctuating pH during fixation and preservation. Buffers are
needed to prevent a reduced pH in either the fixative or pres-
ervation solution to avoid excessive acidity in formalin that
may decalcify bone (Taylor, 1 977). However, tissues clear when
the buffer makes the solution alkaline. Taylor's (1977) data
indicate that pH can fluctuate only in a narrow range without
LAVENBERG ET AL.: PRESERVATION AND CURATION
59
causing some degree of specimen damage. A pH of less than
6.4 begins the process of decalcification, mineral loss in bone,
whereas a pH in excess of 7.0 initiates the clearing process that
results in translucency.
Tucker and Chester (in press) recommend that sodium borate
not be used as a buffer on the basis that it results in high pH,
i.e., loss of pigment may occur. Calcium carbonate also is not
recommended because it tends to precipitate out of solution and
onto the larvae. Hexamine should not be used at all because it
tends to clear specimens independent of pH, and to damage
them (Steedman, 1976)
Markle (1984) summarized five years of data for phosphate
buffered formalin solutions used as a preservative. He used the
standard ichthyoplankton protocol for fixation of his samples.
He gives compelling reasons for using a phosphate buffer to
control pH of formalin solutions used as a preservative for fish
larvae, on the basis that the amount of the buffer can be adjusted
to control pH.
A review of the ichthyoplankton protocols indicates that so-
dium borate (borax) and calcium carbonate (marble chips) are
the preferred buffers, although Tucker and Chester (in press)
recommend sodium acetate. We wish to stress that our knowl-
edge is inadequate, particularly in understanding the chemistry
of these processes. Clearly, a study of the chemistry of fixation
and preservation must occur before a recommendation of an
acceptable buffer can be made. However, we agree with Markle
(1984) that phosphate buffers offer the best alternative to borax
and marble chips for long-term preservation on the basis of
their versatility in adjusting pH.
Presenarives. — Afler the fixation process is completed, the zoo-
plankton collections are processed to obtain data on plankton
volumes. Then the samples are sorted to remove the ichthyo-
plankton component, the eggs and larvae of fishes. After the
identification, enumeration, and measurements offish eggs and
larvae, they are ready for long-term archival preservation.
Through this process the collections are usually maintained in
a buffered formalin solution. However, Ahlstrom (1976) indi-
cated that if an investigator was sensitive to formalin then eth-
anol or a similar preservative was acceptable.
For final long-term archival preservation Ahlstrom (1976)
indicated that fish eggs and larvae were separately vialed, and
placed in fresh preservative. This fresh preservative was a one
percent buffered formalin solution made with freshwater. Ac-
cording to Ahlstrom (1976) the larvae remained in excellent
condition for a period of 15-20 years. Tucker and Chester (in
press) recommend a long-term preservative consisting of a 4%
formalin solution made from distilled water with sodium acetate
used as a buffer. Whenever formalin is used as the basis for a
long-term preservation fluid for fish eggs and larvae there will
be problems of pH. Phosphate buffers apparently control pH
best as they are capable of maintaining pH within a narrow
range between 6.4 and 7.0. Unfortunately the use of formalin
as a final preservative has the potential to incur considerable
curatorial expenses just to monitor pH levels.
We recommend that 70% ethanol be used as the final pres-
ervation fluid on the basis that it renders the pH problem moot,
eliminates working with the fumes of formalin, and eliminates
problems associated with the staining process. In recommending
ethanol we wish to reduce or eliminate the bufliering problems
and their associated pH problems in formalin solutions. After
fixation, the concentration of formalin can be reduced to a 1%
solution, then this fluid can be drained off during the volume
determination process and replaced with ethanol. It is important
to transfer the collections directly from the I % formalin solution
into ethanol without washing them through a water bath. Thus
a small concentration of formalin fixative will be retained in
the ethanol preservative. Also, the transfer should be a staged
one through a series of ethanol solutions, from 1% formalin to
20% ethanol to 45% ethanol to 70% ethanol, rather than a direct
transfer. Zooplankton collections should be stored in the dark,
specifically avoiding light. Also, the storage facility should be
as cold as possible, and it should avoid fluctuating temperatures.
In summary, we recommend that formalin be the fixative of
record until a suitable alternative can be established. Buffers
should be investigated to determine how they affect long-term
effects of fixation and preservation. Phosphate buflfered formalin
is recommended as the most suitable one to control pH within
a narrow range to prevent melanistic pigment loss and deminer-
alization. We recommend that ethanol replace formalin as a
preservative fluid. Finally, the chemistry of fixation and pres-
ervation should be addressed by a chemist to establish a suitable
protocol for processing zooplankton samples.
Curation.— The chief problems with storage and curation of
larval fish collections are to prevent fluid loss, stabilize collec-
tions, and to allow for retrieval availability.
Fluid losses through evaporation in small containers, such as
vials, can be disastrous. There are means to reduce evaporation.
We propose that a double vialing procedure be established (Fig.
27). First, evaporation may be significantly reduced, and second,
a double vialing system provides a mechanism to eliminate
abrasion and damage to fish eggs and larvae. The procedure
calls for an inner vial containing the specimens and preservation
fluid sealed with a poly-seal closure. This vial is inserted into
another glass vial, which leaves sufficient space for labels and
specimen documentation. The second vial is sealed with a plas-
tic closure. The outer vial is placed upside down over the inner
one. The procedure here is to allow gravity to work on vapor
evaporating from the inner vial in such a manner that it must
be compressed before escaping from the outer vial. Essentially
an equilibrium would be achieved that would act to prevent
further evaporation. In addition, a means for specimen docu-
mentation can be achieved that allows for maximizing these
data for curation without causing abrasion or damage to the
delicate specimens.
Another important aspect of this curation technique would
be its contribution to retrieval availability. The vials can be
integrated into an existing ichthyological system so as to make
them immediately available to researchers while offering to
maximize long-term archival preservation protection.
We would like to thank all of our colleagues who provided
us with information relative to the fixation, preservation and
curation of the early life history stages of fishes.
On behalf of the steering committee of the Ahlstrom Sym-
posium we would like to recommend that the National Museum
of Natural History in Washington, D.C., the Museum of Com-
parative Zoology (Harvard University), and the Natural History
Museum of Los Angeles County in Los Angeles be considered
for the deposition of the early life history stages of fishes for
long-term archival care.
Section of Fishes, Natural History Museum of Los Ange-
les County, 900 Exposition Boulevard, Los Angeles,
California 90007.
DEVELOPMENT AND RELATIONSHIPS
Elopiformes: Development
W. J. Richards
THE Elopiformes comprises four genera of recent fishes and
each of these genera is composed of at least two species.
The species are found in tropical waters of the Atlantic, Indian
and Pacific oceans. Elops, a cosmopolitan genus, is composed
of several species and Megalops is composed of two species. M.
atlantica Valenciennes is found in both the eastern and western
Atlantic and M. cyprinoides (Broussonet) is found in the Indian
and western Pacific Oceans. Alhula has two recognized species.
A. vulpes is cosmopolitan and A. nemoptera is found on the
Atlantic and Pacific coasts of the Americas. Recent electropho-
retic work indicates that there may be additional species (Shak-
lee and Tamaru, 1981). Pterothnssus has one species along the
coast of West Africa, P. helloci Cadenat, and one off Japan, P.
gissu Hilgendorf
Larval stages of elopiform fishes have attracted great interest
among ichthyologists because of their unusual leptocephalus
development, a stage found in no other group but the Anguil-
liformes and Notacanthiformes. Consequently most recent clas-
sifications have combined all fish with leptocephalus larvae
into the Elopomorpha (Patterson and Rosen, 1977). Forked tails
of the elopiform leptocephali provide an easy means of sepa-
rating them from other leptocephali which have reduced or no
tails at all. The non-fork tailed leptocephali are treated sepa-
rately in the three subsequent papers in this volume.
Recent classifications have altered our classical view of elo-
piform fishes by suggesting a much closer relationship with eels.
Greenwood et al. (1966) included all fishes with leptocephalus
larvae in the superorder (Elopomorpha). This superorder con-
tained: Elopiformes with two suborders, the Elopoidei (Elopidae
and Megalopidae) and the Albuloidei ( Albulidae including Pter-
othrissidae); Anguilliformes with two suborders, the Anguil-
loidei and Saccopharyngoidei; and Notacanthiformes with two
families (Notacanthidae and Halosauridae). A number of papers
have discussed this proposed classification and a majority has
sustained the opinion that the Elopomorpha is a monophyletic
assemblage. Forey (1973a) discussed the intragroup relation-
ships and made some interesting observations on leptocephali
in a second paper (1973b). Two significant classifications ap-
peared in 1977, one by Greenwood and one by Patterson and
Rosen. Both classifications concluded that Elopomorpha is a
natural, monophyletic group and that Albula and Pterothrissus
are related to the Halosauridae and Notacanthidae. Greenwood
(1977) presented a concept of Elopomorpha as a Cohort Tae-
niopaedia with two superorders: Elopomorpha comprised of
Elops and Megalops in the Order Elopiformes (Suborder Elo-
poidei) and Anguillomorpha comprised of two orders, the Al-
buliformes with two suborders (Albuloidei and Halosauroidei)
and the Anguilliformes. Patterson and Rosen (1977) defined a
cohort Elopomorpha of three orders: Elopiformes, Megalopi-
formes and Anguilliformes, the latter with two suborders— the
Anguilloidei and Albuloidei. Patterson and Rosen (1977) con-
cluded that the interrelationships of the Elopidae, Megalopidae
and Anguilliformes are best represented by an unresolved tri-
chotomy. However, it would seem that those with forked tails
would be monophyletic and the reduced or tailless leptocephali
would be derived from those with tails. The trichotomy scheme
results in paraphyletic forked tailed forms.
With the exception of the species of Pterothrissus. the species
of the remaining genera are coastal with some stages entering
hyposaline environments. Pterothrissus helloci occurs benthi-
cally from 70 to 500 m, most abundantly from 120 to 250 m,
off the coast of West Africa from 9°N latitude to 20°S latitude
(Poll, 1953). All elopiforms are presumed to have pelagic eggs
although the eggs of all are undescribed. According to Smith
and Potthoff (1975) the eggs and early larvae of Harengula
jaguana were erroneously attributed to Megalops atlanticus by
Breder (1944), Mansueti and Hardy (1967), and Mercado and
Ciardelh (1972).
The larval stages have been well described for all genera and
are unique (Fig. 28). The larval stage is represented by the lep-
tocephalus which has been defined by Hulet (1978) and Smith
(1979). The leptocephalus is compressed, transparent and leaf-
like with a mucinous pouch which distinguishes it from all other
fish larvae. It grows to large size compared to other fish larvae,
it has fang-like teeth at the early stages which are subsequently
lost (possibly reabsorbed), its viscera is confined to a narrow
strand along the ventral midline, its musculature forms a thin
layer outside of the mucmous pouch and the remainder of the
pouch consists of a mass of acellular material composed of
mucoproteins and polysaccharides enclosed by a continuous
layer of epithelial cells. Its gut is in two sections, an esophagus
and an intestine which are separated by a gastric region com-
posed of the stomach, liver and gallbladder. The kidney, of
various lengths, lies over the gut beginning near the gastric region
and contmuing posteriorly. Ventral blood vessels conspicuously
appear between the aorta and the kidney and gut. In elopiform
leptocephali dorsal, anal, pectoral and pelvic fins are present
and the caudal fin is large and forked.
Genera of elopiform leptocephali are easily identified except
at small sizes prior to caudal development when myomeres are
difficult to count. The number of myomeres for elopiforms ranges
from 51 to 92 whereas most anguilliform leptocephali have
more than 95. Leptocephali of the Cyemidae have 80 myomeres.
Smith ( 1 979) provides a key, characterizations and illustrations
of the genera. Many other workers have described complete
series or individual stages. Complete series of Elops have been
described by Gehringer (1959a), Megalops by Wade (1962),
Alhula by Alexander (1961), and Pterothrissus by Matsubara
(1942). Among other papers which describe and illustrate var-
ious stages are: oi Megalops by Delsman (1926b), Mercado and
Ciardelli ( 1 972), Gehringer ( 1 959b), Eldred ( 1 967b, 1 972) and
Richards (1969); of Pterothrissus by Smith (1966b) and Rich-
60
RICHARDS: ELOPIFORMES
61
rv' ve- \vN\v V V -^
z^-^^^..
Fig. 28. Elopiform leptocephali. Top to bottom: Elops sp., 33.8 mm SL, Luanda, Angola (redrawn from Richards, 1 969); Megalops allanticus.
22.8 mm SL. Luanda, Angola (redrawn from Richards, 1969): Plerolhnssus belloci. 123.9 mm SL, off Angola (redrawn from Richards, 1969);
and Albula vulpes, 64.2 mm (redrawn from Alexander, 1961).
ards (1969); of Elops by Hildebrand (1963a), Eldred and Lyons
(1966), Gomez Caspar (1981), Richards (1969); and of Albula
by Eldred ( 1 967a), Poll (1953), Gomez Gaspar ( 1 98 1 ) and Hil-
debrand (1963b). The Albula leptocephali heads illustrated by
Meyer-Rochow (1974) may be incorrect.
The characters used for distinguishing the families and genera
(following Smith, 1979) are as follows: Albula and Pterothhssus
leptocephali have the origin of the anal fin well behind the dorsal
fin by a distance exceeding the length of the anal fin base whereas
Elops and Megalops have the origin under the dorsal fin or close
Table 7. Meristic Characters for Selected Elopiform Leptoc ephali.
Taxon
Source
Dorsal rays
Number of anal rays
Myomeres
Elops
saurus
spp.
Gehringer (1959a)
Richards (1969)
21-26 usually 22-24
20
12-15 usually 13-14
15-17
78-82 usually 79-80
70-73
Megalops
allantica
cyprinoides
Wade (1962)
Wade (1962)
9-13 usually 12
10-17 usually 12-17
16-22 usually 19-21
18-25 usually 23-25
51-57
59-68 usually 62-67
Alhula
vulpes
nemoptera
Alexander (1961)
Rivas(1967)
16
7
65-70 usually 67-68
69-74
Pterolhrissus
belloci
Richards (1969)
51-56
10-13
85-92
62
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
behind it, by a distance not exceeding the length of the anal fin
base. Flops and Mega/ops leptocephali have lateral pigment but
Albula and Pterothrissus leptocephali do not have lateral pig-
ment. Elops is distinguished from Mega/ops by having a de-
pressed head, more dorsal than anal rays and the origin of the
anal fin is under the posterior end of the dorsal fin or slightly
behind it. Megalops does not have a depressed head, has fevk'er
dorsal rays than anal rays and the origin of the anal fin is under
the middle of the dorsal fin. Albula leptocephali are separable
from Pterothrissus leptocephali by the distance between the pos-
terior edge of the dorsal fin and the origin of the anal fin. In
Albula this distance is about 2.5 times the length of the dorsal
fin base and in Pterothrissus this distance is about 6-7 times
the length of the dorsal fin base. Also the snout is short in Albula
and prolonged in Pterothrissus. Within genera, meristic char-
acters are useful in identification of the species (Table 7).
The interrelationships of the elopiform fishes are discussed
by Smith in a subsequent paper in this volume.
National Marine Fisheries Service, Southeast Fisheries
Center, 75 Virginia Beach Drive, Miami, Florida 33149.
Notacanthiformes and Anguilliformes: Development
P. H. J. Castle
THE Notacanthiformes (spiny eels) and Anguilliformes (true
eels) were united with the Elopiformes (tenpounders, tar-
pons, bonefishes) by Greenwood et al. (1966) as the superorder
Elopomorpha. These authors noted that members of the three
orders share osteological similarities, swim bladder not con-
nected with ear (except for Megalops), and a distinctive larval
phase (leptocephalus). More recent authors (Nelson, 1973; Fo-
rey, 1973b; Patterson and Rosen, 1977) recognised this rela-
tionship, though not precisely in this form. There seems little
doubt that they are indeed closely related, but in being exclu-
sively elongate fishes the notacanths and eels are readily distin-
guished externally from the short-bodied, herring-like Elopi-
formes.
NOTACANTI form ES
McDowell (1973) reviewed the notacanths, a morphologically
discrete group of fishes, found on or near the bottom on the
deeper continental slope into the deep sea, recognising 2 sub-
orders, 3 families, 6 genera and 22 extant species (Table 8). He
chose to give subordinal distinction to the Halosauridae on the
one hand, and the Notacanthidae and Lipogenyidae jointly on
the other, although Marshall (1962) had already demonstrated
major structural similarities between these families.
The Notacanthiformes have in common with the Anguilli-
formes a leptocephalus phase, an elongate body form, the as-
sociated lengthening of the anal fin, and a reduced caudal fin.
Members of the two orders are otherwise dissimilar. Notacanths
have well developed pelvic fins; a compact, dorsal fin with spines
in some species; scales present and prominent in some; and a
large gill opening and opercular flap. Eels lack pelvic fins; the
dorsal, unless secondarily reduced or lost, is always long and is
supported by delicate rays; scales, if present, are greatly reduced;
and the gill opening and its supporting structures are also re-
duced. Furthermore, notacanth leptocephali are as distinctive
from those of the true eels as are their adults (Fig. 29). They are
greatly elongate (up to 180 cm), having a thin post-caudal fil-
ament in place of a normal caudal fin; dorsal and pelvic fins are
represented by compact, short-based structures present at some
stage of larval growth; there is a minute pectoral, straight gut,
subterminal anus and the myomeres are V-shaped, not W-shaped;
pigment occurs in a ventral series and (rarely) below the mid-
lateral level.
Several quite different notacanth leptocephali of this type are
known, some almost certainly halosaurids ( Tiluropsis. Lepto-
cephalus attcnuatus), some possibly notacanthids (Tilurus) and
others of unknown identity (Leptocephalus giganteus). Eggs and
early larvae have not yet been identified and information on
vertebral numbers is mostly lacking for the group. Until con-
firmed identifications have been made and more information
is forthcoming from leptocephali, ontogeny is unlikely to con-
tribute further to the little that is known of relationships in this
order.
Anguilliformes
The Anguilliformes make up a much larger and more diverse
assemblage. I recognize 21 families. 153 genera and 720 species
for the group (Table 9).
Within the Anguilliformes itself Bohike (1966) reviewed the
Table 8. Composition, Distribution and Habitat of the Nota-
canthiformes. + = All or most species; ( + ) = some species only.
Halo-
saundae
Nola-
canlhidae
Lipo-
genyidae
Taxonomic components:
Known genera (adults)
Known genera (larvae)
Known species (adults)
3
?1
13
2
?l
8
1
0
1
Distribution:
Atlantic: Genera
Species
3
7
2
3
1
1
E. Pacific: Genera
Species
1
2
1
1
0
0
I.-W. Pacific: Genera
Species
2
5
2
4
0
0
Habitat (species):
Shelf
Slope
Abyssal
(+)
(+)
( + )
( + )
( + )
+
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
63
Leptocephalus giganteus 390mm TL
'Tilurus"
"Tiluropsis'
Fig. 29. The three major forms of notacanth leptocephali showing in upper two the elongate snout, distinct dorsal (arrow), and ventral
melanophore series; in lower left the myoseptal pigment; and in lower right the oval eye.
superfamily Saccopharyngoidea (gulpers), a small group of 3
families, 4 genera and 8 species of highly modified mid-water,
oceanic eels, unmistakeable in body form and possessing a lep-
tocephalus of distinctive type. Although they are currently ac-
cepted to be true eels, they are so highly aberrant in form and
osteology that a case could be made for their retention in a
separate suborder, as indeed was proposed by Greenwood et al.
(1966). Other eel families have been studied in some detail,
notably the Congridae (Smith. 1971), Synaphobranchoidea
(Robins and Robins, 1976), Ophichthidae (McCosker, 1977),
Nemichthyidae (Nielsen and Smith, 1978) and others, but there
are several major gaps and the order has never been compre-
hensively reviewed.
With some exceptions, the families and genera of eels occur
worldwide (Table 9) while eel species have a more restricted
distribution in one or other of the major oceans. Some meso-
pelagic, slope/abyssal species and just a few shelf species are
known from both Indo-west Pacific and Atlantic. As for many
other teleosts. the Indo-west Pacific is richest in genera and
species, despite relatively limited collecting there, and infor-
mation is scattered (Alcock. 1889 e/.yf(7!/.: Fowler, 1934;Asano,
1962; Karrer, 1982). The eel fauna of the Atlantic is rather better
known (Blache, 1977; Bohlke, 1978) but by comparison the
group is rather poorly represented in the East Pacific.
Characters.— The families and genera of Anguilliformes are dis-
tinguished principally by external characters, including mor-
phometries (Table 10) but the limits are not yet firmly estab-
lished for all families in the order. Osteological characters, which
mostly reflect these external modifications are also of value at
family and generic levels (Table 1 1 ) but are inadequately known,
especially in the Congridae and related families, and the Mu-
raenidae. Too few genera have been identified in their larval
form for ontogenetic characters to have been used extensively
64
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 9. Composition, Distribution and Habitat of ihe Anguilliformes.
All or most species; ( + ) = some species only.
Synapho- Ophich- Netla- Dench-
branchi- Dysom- Simcn- thi- Con- Muraenc- stomali- Colo- thyi- Semvo- Anguil-
dae matidae chelyidae dae gndae socidae dae congndae dae mendae lidae
Helcr-
Monn- enchelyi-
guidae dae
Taxonomic components:
Known genera (adults)
Known genera (larvae)
Known species (adults)
Distribution:
Atlantic: Genera
Species
E. Pacific: Genera
Species
I.-W. Pacific: Genera
Species
Habitat (species):
Freshwater
Shelf: Tropical
Temperate
Slope/abyssal
Pelagic
3
9
1
55
28
9
6
1
2
3
1
2
2
1
2
0
25
15
4
5
1
2
3
1
2
2
7
16
1
250
131
16
32
4
3
12
15
13
8
3
5
1
29
17
2
6
1
2
2
1
2
2
7
6
1
73
32
5
13
2
3
6
2
2
7
0
1
0
17
10
2
3
0
1
1
0
1
1
0
2
0
39
12
2
3
0
1
1
0
1
1
4
6
1
35
24
7
6
1
2
2
1
1
0
8
8
1
137
( + )
+
( + )
63
+
( + )
9
+
8
( + )
( + )
2
3
6
13
+
10
( + )
+
0
(+)
+
-1-
-1-
+
( + )
( + )
( + )
( + )
-1-
-1-
( + )
( + )
-H
in determining relationships. Eel species are principally distin-
guished externally, by teeth and cephalic pore patterns and by
meristics, especially the number of vertebrae. The latter reflects
the number of myomeres in the leplocephali.
Many of the adult characters by which the families and genera
differ from one another appear to be correlated with the extent
to which the rather sedentary mode of life associated with bur-
rowing, crevice-dwelling or pelagic habits has been elaborated
throughout the group. In most families of eels there are species
in which the body is very slender, with vertebrae numbering
180 or more (Table 10). The pectoral fins are reduced or lost
variously in families (Muraenidae, Heterenchelyidae), genera
(Ophichthidae, Xenocongridae), or even within the life span of
individuals {Moringita). The median fins may also be reduced
to vestiges either in height or in length by a posteriorwards shift
of their origin, or they may be entirely lost, though pterygio-
phores can be retained. Scales occur only in some of the syna-
phobranchoids and in the Anguillidae.
Other characters are not so clearly associated with the adop-
tion of fossorial, cryptic or pelagic habits. These include the
ventral displacement of the gill openings (the extreme devel-
opment being in some Synaphobranchidae and a few Ophich-
thidae where they are confluent ventrally); the ventral displace-
ment of the posterior nostril (most Ophichthidae, Xenocongri-
dae, to some extent the Synaphobranchidae) so that it may even
open within the mouth; or its dorsal displacement (Muraenidae),
Table 10. Some Morphological Characters of the Anguilliformes. + = All or most species; ( + ) = some species only; * = presumed
primitive condition.
Synapho-
Dysom-
Simen-
Ophichlhi-
Con-
Muraenc-
Nella-
Colocon-
Dcr-
branchidae
matidae
chelyidae
dat-
gndae
socidac
stomatidae
gndac
ichthyidae
Vertebrae: Min.*
126
107
121
110
105
120
186
148
126
Max.
172
204
125
270
225
261
290
163
159
Scales: Present*
-t-
( + )
-h
Absent
( + )
-t-
-1-
-1-
-1-
-1-
-1-
+
Pectoral: Present*
+
+
+
( + )
-t-
-1-
Reduced
{ + )
{ + )
(+)
( + )
( + )
-1-
-1-
Absent
( + )
(+)
( + )
( + )
+
Caudal: Present*
-1-
+
+
-1-
+
+
+
H-
Reduced
(+)
(+)
Absent
-1-
Dorsal origin:
Over pectoral/gill opening*
+
+
H-
+
-1-
+
+
+
Between pectoral and anus
(+)
■f
Over or behind anus
(+)
Gill openings: Lateral*
+
+
+
+
-1-
Displaced ventrally
-1-
+
+
(+)
+
Posterior nostril: Before eye*
+
+
+
+
+
+
Displaced dorsally
(+)
(+)
Displaced ventrally
+
+
-1-
-1-
(+)
(+)
Lateral line: Complete*
+
+
+
+
+
-1-
+
+
Incomplete
+
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
65
Table 9. Extended.
Murae-
nidae
Myro-
congn-
dae
Xeno-
congn-
dac
Nem-
ichlhyi-
dae
Cyema-
Iidae
Sacco-
pharyngi-
dae
Eury-
pharyngi-
dac
Mono-
gnathi-
dae
13
1
8
3
2
1
1
2
7
0
6
2
1
1
1
1
170
1
15
8
2
3
1
6
8
1
6
6
2
1
1
2
30
1
7
7
2
3
1
2
6
0
2
3
1
0
1
1
12
0
2
6
1
0
1
2
13
0
7
3
1
0
1
2
120
0
8
7
1
0
1
3
(+)
+
+
+
( + )
+
+
+
+
+
or both (the genera of Nettastomatidae). There may be a re-
duction of the lateral line (Muraenidae. Xenocongridae) or, con-
versely, its great elaboration (the congrid Scalanago). In some
eels there is an enlargement of the mouth and teeth-bearing
surfaces, either by a forward prolongation of the premaxillary-
ethmovomer and dentary (Nemichthyidae, Cyematidae and
others), or by the turning backwards of the suspensorium with
a coincident reduction or loss of the palatopterygoid arch
(Ophichthidae, Muraenidae).
In all eels the branchial region is elongate, the pectoral girdle
is separated from the skull and the posttemporal is lost. This
lengthening is accompanied by a reduction of the opercular
series, narrowing of the gill opening and increased importance
of branchial pump respiration. The branchial series is displaced
backwards with enlargement of the 4th arch as pharyngeal jaws,
especially in the Muraenidae. The long branchial wall is sup-
ported by an increased number of branchiostegal rays which
curve up around the branchial region and expand distally. In
the ophichthids the throat is further supported by numerous
accessory branchiostegal rays (Parr's "jugostegalia") which are
not attached to the hyoid arch and overlap in the ventral mid-
line.
Overall, there is clearly a strong functional correlation be-
tween the lengthening, narrowing and smoothing out of the body
outline, the increase in body flexibility and modifications in
nostrils, jaws, gill openings and lateral line with the mode of
life which is a feature of the eels as a group.
Eggs.—T\\e best known stages in the early life history of the
Anguilliformes (less so in the Notacanthiformes) are undoubt-
edly their highly distinctive leptocephali. Eggs and earliest larvae
are very poorly known. Those of the saccopharyngoids and no-
tacanths have not been identified. Grassi (1913), Schmidt (1913),
D'Ancona(1931b)and Sparta (1937 e^^e^M.) described eggs and
developmental stages of several Mediterranean eel species, mostly
from reared material. The basis for identification of eel eggs was
thus reliably established. Some errors have been made: Eigen-
mann's (1902) eggs of Conger oceanicus were apparently those
of Ophichlhus cruenlifer {Nap\m and Obenchain, 1980); Fish's
(1928) Angiulla rostrala eggs were those of the muraenid An-
archias yoshiae (Eldred, 1968). Little further information has
been added recently, although Naplin and Obenchain's (1980)
detailed account of Ophichlhiis cruent ifer demonsUalcs the use-
fulness of matching planktonic, newly hatched larvae with late
stage embryos. Yamamoto et al. (1975a, b) described live eggs
and early larvae of Angnilla japonica spawned from a ripe fe-
male that had been artificially matured, but there have been few
//; v/vo studies. There is no comprehensive information available
for the identification and comparison of eel eggs, principally
Tabi E 10. Extended.
Serrivo-
Anguil-
Monn-
Heteren-
Muraeni-
Myrocon-
Xenocon-
Nemich-
Cyemati-
Sacco-
Eury-
Mono-
meridae
hdae
guidae
chelyidae
dac
gndae
gridae
ihyidae
dae
pharyngidae
pharyngidae
gnalhidae
137
100
98
108
107
131
97
170
74
138
97
88
170
119
+
180
227
216
131
156
400 +
108
250
125
95
+
+
+
+
+
+
+
+
+
+
+
-1-
+
+
(+)
+
+
+
(+)
(+)
+
+
(+)
(+)
+
+
+
+
-1-
-1-
-1-
+
+
+
-1-
-t-
+
+
+
( + )
+
+
-1-
-1-
-1-
-1-
( + )
+
(+)
+
( + )
+
+
( + )
-1-
+
+
+
+
+
+
-1-
-1-
+
+
+
-1-
-1-
+
+
+
-1-
-1-
-1-
-1-
+
+
+
-1-
+
+
-1-
-1-
+
+
+
-1-
-1-
+
-1-
+
66
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 1 1. Some Osteological Characters of the Anguilliformes. + = All or most species; ( + ) = some species; * = presumed primitive
condition.
Synapho-
Dysom-
Simen-
Ophichthi-
Muraene-
Netlasto-
Colocon-
Dench-
branchidae
matidae
chelyidae
dac
Congridae
socidae
matidac
gndae
thyidae
Frontals: Separate*
Fused
+
+
+
+
+
+
+
+
+
Pterygoid: Present*
+
+
+
+
Reduced
+
(+)
+
+
( + )
Absent
(+)
Hyomandibula: Forward*
+
+
+
+
+
+
Vertical
( + )
(+)
( + )
Backward
+
+
+
Lateral line ossifications:
Present
+
+
+
+
Absent*
+
+
+
+
+
Gill arches:
More or less complete*
+
+
( + )
+
+
+
( + )
Variously reduced
+
( + )
( + )
because only a few species have been studied from just six
families. Major characters of eggs of these families are collated
in Table 12, which also includes selected references. Eggs and
earliest larvae of Ophichthus cruentifer are illustrated as an ex-
ample in Fig. 30.
Eel eggs are large; the chorion is thin and clear, but may have
minute chromatophores; the perivitelline space is wide; the yolk
makes up about one half of the egg diameter and is segmented,
with or without chromatophores. Oil globules are usually pres-
ent (absent in Muraenidae and Nettastomatidae) but the number
and size may vary during development. Development takes
around 4 days at about 20 C in Gnathophis mystax (Thomo-
poulos, 1956) and in O. tTM£'n//7er(NaplinandObenchain, 1980)
but may be several days longer. The yolk reduces in size and
the embryo reaches a hatching length of about 4.5-5.5 mm,
coiling once or more around the yolk. While the late embryo
may possess conspicuous melanophores and segmentation, the
definitive number of myomeres and the characteristic pigmen-
tation of the lai~vae, if any, are not usually fully established until
after hatching.
Leptocephali.—The yolk-sac larva ("preleptocephalus" or en-
gyodontic stage) which is liberated from the egg is characteris-
tically elongate, with a tear-drop shaped to elongate yolk. It
Table 12. Characters of Anguilliform Eggs.
Family
1
2
Ophichthus
Ophichthus
Dalnphi
.
ipterichtus
Ophisurus
Echehis
Ophichthid
Facciolella
Character
cruentifer
remicaudus
imberbis
caecus
.serpens
mvnis
(unident )
oxyrhvncha
Diameter of chorion: Min.
1.62
2.10
2.20
3.00
3,04
3.04
3.40
2.96
Max.
2.89
2.40
2,40
3.60
4.00
3.80
3.68
3.24
Diameter of yolk: Min.
1.32
1.32
1.68
2.10
1.60
1.32
1.48
Max.
1.60
1.60
1.60
1.92
2.20
1.85
1.80
1.84
Oil globule(s): Absent
+
Present
-1-
-1-
-1-
+
-1-
+
+
Number Min.
1
6
1
3
11
1
Max.
1
22
4
40
28
1
11
Size Min.
0.26
0.08
0.32
Max.
0.65
0.16
0.36
0.36
Pigment of embryo:
Present on caudal
-t-
-t-
-1-
+
-1-
Present on gut
-1-
-1-
+
+
+
Present on spmal cord
Chorion smooth:
-1-
-1-
+
+
+
+
+
+
Yolk segmented:
-1-
-1-
+
+
+
+
+
+
Reference
a
b
b
c
d
e
f
g
Families represented:
References:
a-
-Naplin
and Obencham
1980
h — Sparta,
1942a
1 Ophichthidae
b-
-Sparta,
1937
i— Sparta,
1939d
2 Nettastomatidae
c-
-Sparta,
1938a
j— Sparta,
1939b
3 Xenocongridae
d-
-Sparta,
1939c
k — Sparta,
1938b
4 Congridae
e-
-Sparta,
1940a
1— Castle and Roberison,
1974
5 Muraenidae
f-
-Sparta,
1940b
m — Mannaro, 1971
6 Anguillidae
g-
-Sanzo,
1938a
n-Eldred,
1969
0— Yevseyenko, 1974
CASTLE: NOTACANTHIFORMES. ANGUILLIFORMES
67
Table 11, Extended.
Scmvo-
mendae
Anguil-
hdac
Monn-
guidae
+
+
+
+
+
+
+
+
+
Heleren- Myrocon- Xenocon-
chelyidae Muraenidae gndae gridae
Eury-
Ncmich- Sacco- pharyngj- Mono-
thyidae Cyematidae pharyngidae dae gnathidae
+?
+?
+?
(+)
+
+?
+
+
+?
somewhat resembles later stages but the development of larval
characters is progressive. There may be substantial differences
in pigmentation between this stage and the fully grown lepto-
cephalus (e.g.. the congrid Ariosoma, Table 17 E,, Mj, O and
Fig. 37); typically the pigmentation pattern is much less com-
plex. The engyodontic stage has few, needle-like teeth, lower
jaws equal to, or longer than upper, an unformed nasal capsule,
and undifferentiated median fin-folds and hypurals.
At about 20 mm TL the leptocephalus then enters the eury-
odontic stage which lasts until metamorphosis. It begins with
shedding of the engyodontic teeth and their replacement by 3
series (usually) of shorter, broad-based teeth, the lower jaw
shortens relative to the upper, the head decreases in relative
length, and the fins and hypurals differentiate. At this stage
leptocephali are highly distinctive and well-known forms amongst
fish larvae. At full growth they are typically around 50-80 mm
but may attain 300-400 mm (Nemichthyidae) or 1,800 mm
(Notacanthiformes). They are almost transparent except for eye
and other pigmentation and the blood lacks erythrocytes and
haemoglobin. The body is greatly compressed and leaf-shaped
or filamentous, typically with a small head, prominent, for-
wardly-directed larval teeth and a posteriorly placed anus. The
electrolyte make-up of their body fluids differs markedly from
that of postmetamorphic forms (Hulet, 1978).
Table 12. Extended.
Family
2
3
4
5
6
Neltastoma
nielanurum
C'h/opsis
hicolor
Conger
conger''
Ariosoma
baleancum
(jnalhophis
Gnathophis sp, mystax
Muraena
helena
(ivmnothorax
unicolor
( i nigro-
marginanis
Angiulla
iinguiUa?
2.40
3.00
1.44
1.48
+
2.72
3.04
1.40
1.48
-I-
13
0.04
0.08
2.60
1.7
-I-
1
0.40
1.80
1.92
1.00
1.04
-I-
1
5
0.30
2.93
3.43
1.25
1.50
-I-
1
9
0.03
0.10
2.50
3.00
1.50
1.85
-t-
5.0
5.5
2.3
3.4
3.3
4.0
1.5
2.0
-I-
2.3
2.9
1.3
1.6
1
2
0.31
0.42
-I-
+
J
+
+
+
+
1
+
+
m
+
-I-
m
-I-
-I-
-I-
n
68
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
B C
mm
D
OPHICHTHUS CRUENTIFER
Fig. 30. Embryonic and early engyodontic stages of Ophichthus cruenltfer (adapted from Naplin and Obenchain, 1980).
I
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
69
Metamorphosis follows the euryodontic stage. It is relatively
abrupt and involves the replacement of many of the character-
istic leptocephalus features by those of the juvenile. The body
rounds up in section, tissue transparency is lost, the postorbital
portion of the head lengthens, the larval teeth are lost and the
definitive teeth are gradually substituted. The anus and median
fin origins move forwards, though not in all species. Pectoral
and caudal fins are lost late in metamorphosis in those species
which lack the fin in the juvenile and adult. There may be a
substantial reduction in body length, extremely so in the No-
tacanthiformes. The principal characters which are retained are
the definitive number of myomeres/vertebrae which is estab-
lished very early in larval life, the number of dorsal and anal
fin-rays which is attained rather late in development, and for
some species the larval pigment. The maintenance of larval
pigment through metamorphosis is of prime importance in iden-
tification at the generic level. However, metamorphic larvae are
relatively rare in collections, possibly because they are in any
case a transient stage; metamorphics are also benthic and hence
less accessible to collection. Information on these important
stages is therefore sparse.
Identification
Leptocephali are thus readily recognisable amongst other fish
larvae, apparently abundant in the warmer ocean, and accessible
near the surface. Large collections of leptocephali have accu-
mulated, for some families and genera there being many more
specimens available than of the adults (e.g., the moringuid, Neo-
conger. Smith and Castle, 1972; the Nettastomatidae, Smith
and Castle, 1982). The availability of such collections and the
need for identification of leptocephali have resulted in the recent
rapid advance of larval studies (Castle. 1969; Blache. 1977;
Smith. 1979; Fahay. 1 983). These studies have, understandably,
emphasized identification rather than inter-relationships based
on larval characters.
Larvae of all but the monotypic families Simenchelyidae and
Myrocongridae and those of about half (82) of the genera are
known. Several distinctive larval forms, possibly of undescribed
genera rather than families, are also known (e.g., the congrid-
like Leptocephalus thorianus Schmidt, Smith, 1979). Family
identification, largely by morphological and pigment characters,
may be arrived at from Table 13, which incorporates infor-
mation set out in key form by Smith (1979) and Fahay (1983).
This "look-alike" approach to identifying leptocephali largely
suffices at the family level but is less satisfactory in identifying
genera, especially of the Ophichthidae and Congridae (Leiby,
1981). More detailed information may be necessary, especially
for species identification, but this will be slow to accumulate.
Some attempt to collate available data for identification pur-
poses is made in Tables 14-23, with their complementary figures
(Figs. 34 to 43).
More than 500 different leptocephali have been described,
200 as nominal species of the invalid genus Leptocephalus Gron-
ovius, 1763. The procedure of formally naming eel larvae in
this way has been both opposed (Bohike and Smith, 1968) and
advocated (Castle, 1969). However, nomenclatural problems
associated with naming larval forms will not be readily over-
come by ignoring the priority of larval names or attempting to
apply a blanket restriction on their use. Some alternative ref-
erence scheme, or at least an agreed descriptive procedure, does
seem appropriate (Fahay and Obenchain, 1978) to accommo-
date the large number of distinctive ontogenetic stages of eels.
Fig. 31. Anterior region of leptocephalus of an unidentified ?net-
tastomatid (DANA St. 4181 II, 34<'23'N, 25°53'W, 9 June 1931), show-
ing tab-like extensions of the intestine.
Few complete growth series have been described and illus-
trated, and developmental osteology is known only for Anguilla
anguilla (Norman, 1926b), Serrivoiner spp. (Bauchot. 1959).
Ariosorna baleancum (Hulet. 1977). Ophichthus gomesi (Leiby,
1979a), and Atyrophispunctatus {Leiby, 1979b). At least in Oph-
ichthus gomesi ossification of the head skeleton does not occur
for most elements until metamorphosis, although the jaws, sus-
pensorium and branchial skeleton are present as cartilage during
the pre-metamorphic stage. Leiby's recent papers (1979b, 1981)
contain detailed information on the sequence of development
of the skeleton and emphasize the relevance of a more thorough
evaluation of developmental osteology in identification of lep-
tocephali.
In overall body form leptocephali range from the greatly elon-
gate notacanths (Castle, 1973, for references; Smith, 1979; Fig.
29), Nemichthys (Nielsen and Smith, 1978; Smith, 1979; Table
19) and some Nettastomatidae (Smith and Castle, 1982) to the
short, deep larvae of Thalassenchelys (Castle and Raju, 1975;
Table 22 and Fig. 42). the Xenocongridac (Smith. 1969; Table
22 and Fig. 42) and Cyema atrum (Smith, 1979; Table 23 and
Fig. 43).
The snout is typically rather sharp, especially so in some
Notacanthiformes (Fig. 29), Dysommatidae (Table 14 and Fig.
70
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
20 3 mm
ENGYODONTIC
37'» mm
Gnathophis
856 mm
EURYODONTIC
Fig. 32. Development of teeth-series in the congrid Gnathophis.
34), Nettastomatidae (Table 19 and Fig. 39) and Cyematidae
(Table 23 and Fig. 43), but characteristically short and rounded
in the Heterenchelyidae (Table 18 and Fig. 38) and Muraenidae
(Table 21 and Fig. 41), especially near metamorphosis. In some
Dysommatidae (Table 14 and Fig. 34) it is produced forwards
as a conspicuous, narrow, ethmoid rostrum bearing at its tip a
pair of "premaxillary" teeth and, in some also, fleshy tabs or
tentacles along its length. The rostrum itself is lost at meta-
morphosis so that the snouts of post-metamorphic dysomma-
tids, apart from their characteristic papillae and plicae, are sim-
ilar to those of other eels.
In full-grown leptocephali the anus lies just in advance of the
midpoint (some Nettastomatidae, Table 19 and Fig. 39; some
Muraenidae, Table 21 and Fig. 41; some Xenocongridae, Table
22 and Fig. 42), well behind the midpoint (most genera), or is
subterminal (the congrid group Ariosoma-Bathymyrus. Table
1 7 and Fig. 37). For those in which it is subterminal, it advances
during metamorphosis, taking with it the anal fin origin and the
developing pterygiophores and actinotrichia. Its position in these
species is thus a very rough measure of the stage of metamor-
phosis. Broadly speaking, the amount of forward movement of
the anus is correlated with the length of larval life, generally
long in Notacanthiformes, Anguillidae (1-3 years) and Congri-
dae (10 months for species of Gnathophis, Castle, 1968; Castle
and Robertson, 1974) but much shorter in Moringuidae (3'/2
months for Moringua edwardsi. Castle, 1979) and probably also
for Muraenidae, Xenocongridae and many Ophichthidae. How-
ever, little is known of the duration of larval life in most eels.
A special feature of some Ariosoma-Bathymyrus larvae is an
exterilium or external intestine (Mochioka et a!., 1982; Table
17Q and Fig. 37) and in the unidentified larva illustrated by
Weber (1913) and Smith (1979), there are tab-like extensions
of the intestine, of unknown significance (Fig. 31).
The olfactory organ is a round to oval sac immediately in
front of the eye. As growth proceeds its single aperture pro-
gressively becomes vertically subdivided by flaps growing from
the upper and lower margins. After separation of the two nos-
trils, the olfactory sac lengthens in many leptocephali, except
the Cyematidae, Nemichthyidae and Serrivomeridae, so that
the anterior nostril moves forwards to near the tip of the snout.
There it becomes subtubular and often turns downwards; late
in metamorphosis the posterior nostril may move dorsally or
ventrally to adopt its final position above or behind the eye or
ventrally on or through the upper lip.
The eye is usually round, but in the notacanthiform larvae
referred to the larval genus Tiluropsis, and in Leptocephaliis
attemiatus, it is characteristically oval, with the long axis ver-
tical. In all Synaphobranchoidea, probably also including the
Simenchelyidae, the eye assumes a so-called "telescopic" or
"tubular" shape (Table 14 and Fig. 34) and the body of the eye
faces anterodorsally and is elongate, with a very deep retina.
Teeth develop shortly after hatching. These engyodontic teeth
(Fig. 32) are few, needle-like, forwardly directed, each one pro-
gressively shorter along the rami of the jaws; typically there is
a pair of larger teeth anteriorly. The engyodontic teeth are shed
at the beginning of the euryodontic growth stage and are pro-
gressively replaced with the 3 series of shorter, broad-based teeth
in upper and lower jaws; the upper teeth are preceded by an
anteriormost pair, slightly smaller than the first maxillary pair,
which are very large in the supposed xenocongrid Thalassenche-
lys (Table 22 and Fig. 42). As growth proceeds teeth are added
progressively, to reach 40-50 at metamorphosis. They are blade-
like and slightly recurved in Paraconger, bicuspid in Coloconger
(Table 1 8 and Fig. 38), or needle-like and distinctly spaced in
the Heterenchelyidae (Table 18 and Fig. 38). Leiby (1979b)
notes that the splanchnocranium is so weakly developed in the
engyodontic stage of the ophichthid Myrophis pimctatus that
the first series of larval teeth cannot be used in feeding.
I
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
71
150
UO
130-
120-
nO'
CO
O
O
^ 100
c/)
O
I
CO
5
90
70'
60
50
40
MYROPHINAE
1 Myrophis punctalus
2 M. plumbeus
3 Ahlia egmontis
u Pseudomyrophis nimius
OPHICHTHINI
5 Aplatophis chauliodus
6 Ophichthus rex
7 O. ocellalus
8 O. gomesi
9 O. cruentifer
10 O. melanoporus
11 Echiophis mordax
12 Myrichthys oculatus
13 M. acuminatus
SPHAGEBRANCHINI
u Ichthyapus ophloneus
15 Apterichtus ansp
16 ^. kendalh
17 Stictorhinus potamius
CALLECHELYINI
18 Letharchus velifer
19 Callechelys muraena
20 C, springer!
21 C. perryae
BASCANICHTHYINI
22 Carolophia loxochila
23 Bascanichthys scuticans
2'- 6. bascanium
25 Gordiichthys irrelitus
26 Phaenomonas longissimus
/»
/•
25
-y^'
i^^
A^^
20
.^^^
^^,^^'
-\^^
^^
'23
^i' 019
022
04
017
150 7-
1=
/
016
•3
1
•^
13
09
^^
oio
02
cTu
"-r~
110
I
120
I
130
I
uo
— I —
150
— I —
180
100
I
160
1
170
190
— I —
200
— I —
210
— I —
220
MEAN TOTAL VERTEBRAEIADULTS) MYOMERESO-ARVAE)
Fig. 33. Position of kidney in adults and larvae of 26 species of Western Atlantic Ophichthidae; black circles adults, open circles larvae.
Adults of not all species shown.
The gill opening is anteroventral to the pectoral base and any
movement to take up an adult ventral position (Synaphobran-
choidea, Ophichthidae) does not occur until very late in meta-
morphosis.
Pectoral fins are present as fleshy tabs in all very early lep-
tocephali. If absent or much reduced in the post-metamorphic
stage, the loss does not occur until late in larval life or at meta-
morphosis (Muraenidae, Ophichthidae, the muraenesocid Gav-
laliccps). Actinotrichia do not develop until late in the eury-
odontic stage and lepidotrichia not until metamorphosis. The
range is 8-22 among the species of eels.
Median fins are first visible as undtflferentiated folds of tissue
and remain so until the beginning of the euryodontic stage. The
dorsal and anal fin skeletons begin to develop posteriorly first,
and then progressively forwards, the anal more rapidly than the
dorsal. Pterygiophores and associated muscle blocks appear be-
fore the actinotrichia but lepidotrichia do not complete devel-
opment until metamorphosis is complete. The anal fin supports
are usually closely packed before the anus moves forwards dur-
ing metamorphosis. The dorsal origin is less easy to define until
late in the euryodontic stage and may not take up its final po-
sition until well into metamorphosis. In the muraenids Anar-
72
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 13. Major Morphological and Pigment Characters of Anguilliform Leptocephali (Families). + = All or most species; ( + ) = some
species only.
Synapho-
Dysom-
Simen-
Ophichthi-
Muraeneso-
Nettasto-
Colocon-
Derich-
branchidae
matidac
chelyidae
dac
Congndae
cidae
matidae
gridae
thyidae
Eye: Tubular
+
+
?+
Normal
+
+
+
+
+
+
Hyomandibula: Backwardly oblique
Normal
+
+
? +
+
+
+
+
+
+
Gut: A simple, straight tube
+
+
+
+
+
with swellings or loops
1 Swelling
2 Swellings
(+)
3 Or more
+
+
( + )
(+)
Body depth: a50%TL
Much <50%TL
+
+
+
+
+
+
+
+
Tail tip: Broad, rounded
Narrow
+
+
+
+
+
+
+
+
Gut length: sHalfTL
( + )
( + )
>HalfTL
+
+
+
+
+
( + )
+
+
Head: Elongate
( + )
( + )
Short
+
+
+
( + )
+
( + )
+
+
Snout: Rounded
Acute
+
+
+
+
+
+
+
+
Pigment: Entirely absent
At least some present
+
+
+
+
+
+
+
+
None on gut
+
+
Present on gut
+
+
+
+
+
+
Present dorsally in orbit
( + )
Absent from orbit
+
+
+
+
+
+
+
( + )
Present on spinal cord
( + )
Absent from spinal cord
+
+
+
+
+
+
+
+
Patch below iris
+
Absent below iris
+
+
+
( + )
+
+
+
+
chias. Uropterygius and to a lesser extent Channotnuraena the
dorsal and anal fins are much restricted and distinctive as such
early in the euryodontic stage (Table 21 and Fig. 41). At least
in the Ophichthidae (Leiby, 1982), even in those species which
lack a dorsal fin in the adult, pterygiophores and actinotrichia
develop in the larvae. There is also a marked correlation between
position of dorsal fin origin in larvae and adults. In the congrid
Ariosoma and related genera, the anus is subterminal and the
dorsal and anal are also restricted but develop progressively
forwards during late larval growth (Table 17 and Fig. 37). Dorsal
fin-rays range in number from 1 10 in Neocyema erythrosoma
to 600-700 in some ophichthids, anal rays usually being some-
what fewer. The large number and apparent considerable vari-
ability of median fin rays in most eels has resulted in this meristic
character being neglected, but it may be of considerable use in
larval identification (Leiby, 1981).
The caudal fin develops at least as early as the anal, its sup-
porting structure being 3 hypurals, the first two joined distally,
enclosing a foramen. Typically hypurals 1 and 2 support 4 rays,
hypural 3 supports 5 rays, but the hypurals are much broader
in the Synaphobranchoidea, supporting about 16 rays. The fin
is resorbed, the rays shorten, and finally become embedded in
the tail tip of heterocongrin and many ophichthid larvae shortly
before metamorphosis.
Myomeres differentiate during embryonic development but
because of their relatively high number and small size it is not
known for any species whether the definitive number of the
adult is established then, or after hatching. However, differen-
tiation of the most posterior myomeres, as evidenced visually,
appears to occur during the engyodontic stage, even for species
with very high total numbers of myomeres. Total counts for
species with more than about 180 are difficult to make accu-
rately, even in fully grown leptocephali. Myomeres are less readily
counted as body transparency is lost at metamorphosis. The
range in myomere number across the Anguilliformes is 74-78
in the short-bodied Cyema atrum to more than 400 in the greatly
elongate Neinichthys scolopaceus (Table 10) with ranges for
species of about 10 myomeres at the lower end (e.g.. for Anguilla,
Jespersen, 1942) to about 30 in the range 200-300 (e.g., for
Nettastomatidae, Smith and Castle, 1982).
Vertebrae first begin to differentiate posteriorly just before
metamorphosis with the constriction of the terminal portion of
the notochord proceeding anteriorly.
The value of vertebral counts in defining eel species has be-
come firmly established in eel studies (Bohlke, 1978). The cor-
relation of vertebral number with number of myomeres in larvae
was demonstrated by Jespersen (1942) for Angidlla and taken
upextensively in recent years (Blache, 1977; Smith, 1979; Smith
and Castle, 1982). In utilizing this agreement between larvae
and adults, associated phenomena need to be further explored
and assessed, e.g.. pleomerism (the correlation in related species
of vertebral number and maximum body length attained: Lind-
sey, 1975), "Jordan's Rule" (the tendency for fishes in polar or
cool waters to have more vertebrae or other meristic parts than
have related forms in tropical warm waters, Jordan. 1892), and
sexual dimorphism in vertebral number (as occurs in Aforingua
edwardsi. Castle and Bohlke. 1976).
The existence of latitudinal dines in vertebral number in eels
has been proposed, but not convincingly demonstrated, except
possibly for the muraenid Gymnothorax panamensis which
CASTLE: NOTACANTHIFORMES. ANGUILLIFORMES
Table 13. Extended.
73
Scrrivo-
mcndae
Anguil-
lidac
Monn-
guidac
Heicrcn- Myrocon-
chelyidae Muraenidae gndac
Xenocon-
gndae
Nemich- Sacco- Eury- Mono-
ihyidae Cyematidae pharyngidae pharyngidae gnathidae
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
{+)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
(+)
(+)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Randall and McCosker (1975) show to have a mean vertebral
range of 1 43 in Chile and 1 25 in the Gulf of California. Variation
across longitude is apparently not usual but may be consider-
able; for example, McCosker (1977, 1979) demonstrates that
the ophichthid Myrichlhys maculatus has a mean vertebral count
of 153 in the East Pacific to 195 in the Red Sea.
Two other problems arise in using vertebral/myomere char-
acters in matching leptocephali with their adult species. These
are the prevalence of damaged tails in adults of some species,
especially those that are slender-tailed (Nettastomatidae, some
Congridae and Muraenesocidae) and hence the unavailability
of vertebral counts; and the overlap or near concordance of
vertebral numbers within species groups. For example, in the
western Indian Ocean there are 15-20 species of the muraenid
genus (iymnolhorax which have vertebral numbers within the
range 130-145. Unless other characters (e.g., fin-ray numbers)
can be shown to differ significantly between these species, it is
likely that their leptocephali, all having rather similar pigmen-
tation, will prove difficult, if not impossible, to identify.
However, there is a reliable correlation between the segmental
position of the larval kidney and that of the adult. The larval
nephros (opisthonephros) is typically an elongate sac lying above
the gut approximately in the middle of the body, i.e., near the
anus in those larvae with a relatively short gut (Xenocongridae,
Nettastomatidae, Ophichthidae) or some distance in front of it
in those having a long gut (Congridae). The segmental position
of the kidney changes little, ifat all, during larval life and through
metamorphosis into the juvenile. Its position then very ap-
proximately agrees with the end of the body cavity and the first
caudal vertebra. The correlation in the nephros position has
been successfully employed as an identification character for the
Muraenidae and other families (Blache, 1977) and for some
Ophichthidae (Leiby , 1981) but its value has not yet been com-
prehensively explored across the Anguilliformes as a whole.
Further evidence for the stability of nephros position from larva
to adult, at least in the Ophichthidae, is provided in Fig. 33.
The figure expresses the mean segmental positions of the end
of the nephros in the larvae and adults of various western At-
lantic ophichthids of the subfamily Myrophinae and the four
tribes of the subfamily Ophichthinae. There is close agreement
in position of the nephros between larvae and adults of all
species. Furthermore, the position of the kidney (and first caudal
vertebra) is conspicuously further back along the body in the
tribes Callechelyini and Bascanichthyini. These are readily re-
cognisable short-tailed ophichthids whose larvae can be im-
mediately identified as such by the posterior position of the
nephros. There is considerable overlap in this character between
the Myrophinae, Sphagebranchini and Ophichthinae although
individually the species are distinct.
The larval nephros is typically supplied and drained by two
prominent blood vessels passing vertically between the lateral
muscles to the aorta and cardinal veins below the vertebral
column. The segmental position of the last of these vessels in
the leptocephalus and its correlation with the position of the
first caudal vertebra in the adult has been emphasised in larval
identification. However, it seems simpler to use nephros posi-
tion instead.
In those groups of larvae in which the anus does not move
forwards during metamorphosis, there is some agreement be-
tween number of preanal myomeres and preanal vertebrae.
74
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 14.
Pigment and Morphological Characters of the Synaphobrachoidea.
to Fig. 34.
+ = All or most species; ( + ) = some species only. Refer
atei
al pigment
A.
A large midlateral patch
at about level of anus
B.
On caudal only
C.
A midlateral row of compact
or dendritic spots
1 . Row complete
2. Postanal row only
D.
A dorsolateral row
E.
A ventrolateral row
F.
A ventral row
G.
A postanal row
Gut pigment
H. Absent
1. An irregular series of dendritic
melanophores along its length
Morphological
J. Posterior flexures of myomeres
rounded
An opaque midlateral area of
myomeres along length of body
Posterior flexures of myomeres
angular
Rostrum absent
Rostrum present
Gut straight
Gut swollen or lightly arched
at points along its length
Posterior end of gut markedly
flexed downwards
K.
M.
N.
O.
P.
Taxa
Synapho-
bronchus
Nettodarus
Dvsommma
Type
Characters
A
B
C
D
(+)
+
+
+
+
+
+
+
+
+
(+)
+
+
+
+
(+)
(+)
(+)
(+)
(+)
(+)
+
+
+
(+)
(+)
(+)
(+)
(+)
(+)
(+)
+
+
+
+
However, this character is not generally applicable in larval
identification because of forward movement of the anus during
metamorphosis in some species.
The gut is most often a narrow straight tube, flexed down-
wards under the pectoral fin and following the ventral margin
to the posteriorly placed anus. The stomach is usually visible
as a finger-like sac at about segment 10. The most frequent
modifications of the gut tube are loops or swellings at intervals
along its length, each usually accompanied by groups of mela-
nophores (Ophichthidae, Tables 15-16 and Figs. 35, 36; Ac-
romycter. Table 18 and Fig. 38; some Nettastomatidae, Table
19 and Fig. 39). The number and state of development (low,
moderate or conspicuous) of the swellings may be diagnostic at
family, genus or species level but is not always so (Leiby, 1981).
The liver, with associated gall bladder, fills much of the space
anteriorly between the gut and the ventral margin of the lateral
muscles. It has two or three lobes in the Ophichthidae (Table
15 and Fig. 35), the gall bladder on the second or third lobe,
and the lobes may be distinct or connected by a thin band of
liver tissue.
Larval pigment is present in larvae of all families except the
Anguillidae and may be highly elaborated to form complex and
distinctive patterns. The pigmentation, if present, is usually much
simpler in the engyodontic stage than later stages. Melanophores
may begin to appear in the embryo (in some Ophichthidae as
several pigment patches on the gut similar to those in the larvae;
in some Muraenidae on the spinal cord) but typically do not do
so until the early engyodontic stage. Pigmentation sometimes
reaches its full expression by the beginning of the euryodontic
stage but typically the complex patterns characteristic of the
Ophichthidae and other families are not complete until full
larval growth. Subsequently pigment may be lost during meta-
morphosis (the congrid Ahosoma), but may serve as a highly
important character in matching larvae with adults.
Individually, melanophores may be dendritic (Dysommati-
dae. Table 14 Ci-C, and Fig. 34). ocellate (Congridae, Table
18B and Fig. 38B), compact (Muraenidae, Table 21 D and Fig.
4 1 ) or rather diffuse (Moringuidae, Table 23 C, and Fig. 43). They
may be isolated, grouped in clusters to form conspicuous pig-
ment patches (the congrid Bathymynis. Table 1 7G and Fig. 37),
or they may form well defined lines, series or patterns. In most
families they occur on the lateral body surface, including the
caudal fin, on the myosepta (Ariosoma, Table 17E and Fig. 37;
Bathymyrus. Table 1 7E and Fig. 37; many Ophichthidae, Table
16 and Fig. 36), or on the ventral body wall (Dysommatidae,
Table 141 and Fig. 34; Congridae, Table 18Land Fig. 38). They
may occur deeper in the tissues, either on the gut, liver, kidney,
suspended in the mucinous space between the lateral muscles,
associated with the spinal cord or vertebral column or, fre-
quently, on the bases of the caudal, anal and dorsal fin-rays.
Although Blache (1977) and Fahay and Obenchain (1978)
have attempted to summarise pigment patterns in some groups
CASTLE: NOTACANTH I FORMES, ANGUILLIFORMES
75
Fig. 34. Illustrations accompanying Table 14.
76
ONTOGENY AND SYSTEM ATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 15. Morphological Characters of Ophichthidae (Myrophinae and Ophichthinae). + = All or most species; ( + ) = some species
only. Refer to Fig. 35.
Myrophinae
Ophichthinae
Characters
Murae- Neen- Pseudo- Ophich- Sphage- Bascanich- Calle-
4h!ia mchlhys Myrophis chetys myrophts thini branchini thyini chelyini
A. Body depth (euryodontic stage)
1. >10%TL
2. <10%TL
B. Gut loops or swellings
1. Low
2. Moderate to pronounced
C. End of nephros
1. Above or just before anus
2. 4-14 myomeres before anus
D. Liver lobes and oesophageal swellings
1. Two
2. Three
E. Caudal fin at metamorphosis
1. Present, normal
2. Absent (or much reduced)
F. Dorsal pterygiophores and rays before
metamorphosis
1 . Well developed; dorsal origin
migrates forwards 4-6 myomeres
2. Weakly developed; origin migrates
forwards 5-50 myomeres (or resorbed)
+ (+)
+ (+) +
(+)
+
(+)
+
(+)
+
(+)
(+)
+
(+)
(+)
of lai-vae, the significance of these has not yet been comprehen-
sively reviewed across the Anguiliiformes. Furthermore, the ex-
tent of intraspecific variability of pigment patterns has also not
been assessed. Any present discussion as to the significance or
otherwise of similarities and differences in larval pigmentation
must therefore be preliminary.
The range of pigmentation in genera for which larvae have
been identified, and for some other forms, is summarized in
Tables 14-23, family by family. These tables, with their accom-
panying figures and morphological information, may be used
as a guide to generic identification, and also as a synopsis of
pigment patterns. Because these are both complex and diverse
in some families, they cannot always be simply displayed in
keys. In the Ophichthidae also, and other families, further pig-
ment patterns are known, probably representing other genera.
This is particularly so of Indo-Pacific Anguiliiformes which have
not been extensively studied.
These tables and figures highlight common features of pig-
mentation: (1) on the gut or its adjacent body wall, often as a
regular, spaced series from throat to anus (Notacanthiformes,
Congrinae, Heterocongrinae. Heterenchelyidae, Colocongri-
dae), or as an interrupted series (Nettastomatidae. Muraene-
socidae. Dysommatidae. Ophichthidae) or in some other form
(Bathymyrinae, Heterocongrinae, Muraenidae, Nemichthyidae,
Xenocongridae); (2) on the lateral body surface (Dysommatidae,
Congrinae, Nettastomatidae, Xenocongridae). often associated
in some way with the myosepta (Ophichthidae, Bathymyrinae,
Heterocongrinae, Serrivomeridae. Derichthyidae); (3) on the
spinal cord (Nemichthyidae. Muraenidae); or (4) on the bases
of the dorsal, anal and caudal fins.
The broad perspective on the ontogeny of the Anguiliiformes
and Notacanthiformes given by the preceding deserves com-
ment.
As adults, eels have adopted a somewhat conformist body
plan notable for reduction and loss of external features, though
the component families of the group are more or less discrete
osteologically. In contrast, through elaboration of the leaflike
body form and pigment patterns their larvae display a diversity
which matches that of any other group of teleosts. This diversity
involves some distinctive larval characters (morphological and
pigmentary) which allow leptocephali to be identified at the
family level. These characters have not been comprehensively
assessed; further definitive identification of larval forms will aid
any future analysis. Within families, larvae are generally similar
in body form and pigmentation but there are several remarkable
exceptions. There are some discernible character gradients in
larvae (e.g., the complexity of gut swellings or loops in Ophich-
thidae; pigmentation of Congridae). but these may or may not
be matched by adult character gradients. Detailed meristic in-
formation, as forthcoming throughout larval development, is
the only satisfactory medium for species identification, espe-
cially in the larger eel families.
Zoology Department, Victoria University of Wellington,
Wellington, New Zealand.
CASTLE: NOTACANTHIFORMES. ANGUILLIFORMES
77
OPHICHTHINAE
Fig. 35. Illustrations accompanying Table 15.
78
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 16. Pigment Characters of Ophkhthidae (Mvrophinae and Ophichthinae).
to Fig. 36.
+ = All or most species; ( + ) = some species only. Refer
Characters
Myrophinae
Ophichlhinae
Bas-
Aturae- Myr- Seen- Pseudo- Ophich- Sphagc- canich- Calle-
Ahiia mchthys ophis chelys myrophis thini branchini Ihyini chelyini
B.
C.
D.
E.
F.
Lateral pigment
A. Absent
A single spot mid-laterally on nearly
every myomere
An oblique row (or streak) of compact spots
below midlateral level
1. On all or most myosepta
2. On only a few myosepta, often associated
with deep axial pigment
Round groups of spots scattered over body
Extra spots on dorsal and ventral myosepta
A group of spots midway along body
Axial pigment
G. Several deep postanal pigment clusters below
vertebral column (sometimes preanal also;
may be associated with myomere pigment)
Gut pigment
H. Scattered spots along gut, usually prominent
groups above upward loops, below downward
loops
Irregular along length, mostly between nephric
duct and crest of each gut loop
Loop pigment associated with spots on body wall
Conspicuous pigment patch at crest of each gut loop
I.
J.
K.
( + )
( + )
(+) +
( + )
( + )
(+)
(+)
( + )
( + )
( + ) +
(+)
{+)
(+)
+
(+)
(+)
+
Head pigment
L. Spots along upper jaw near bases of teeth and
often on lower jaw
M. On postorbital region, pectoral base or
oesophagus
Other pigment
N. On bases of anal rays
O. On body wall above anal base
P. On bases of dorsal rays
Q. On body wall below dorsal base, or before it
R. On caudal base
+
(+)
(+)
+
+
+
+
+
(+)
+
(+)
+
+
+
+
(+)
+
+
+
(+)
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
79
OPHICHTHIDAE
Fig. 36. Illustrations accompanying Table 16.
80 ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 1 7. Pigment and Morphological Characters of Congridae (Bathvm'i rinae, Heterocongrinae) and Miiraenesocidae. + = All or
most species; ( + ) = some species only. Refer to Fig. 37.
I^ara-
Allo- Ario- Uniden- Balhv- Paru- Goi- fivlern- Con- (iuvi- Miirae- xcnonjy- .\cnn-
conger soma tified niyriis conger gasta conger gresox aliccps ncso.x s/av mv^lax
Lateral pigment
A. Absent
B. A midlateral row of single spots,
often with extra spots below
C. A row of few large spots between
midlateral and ventral levels
D. A large group of dendritic spots
at about myomere 80
E. Oblique rows of compact spots on
myosepta below midlateral level
1 . Spots very close together + + + +
2. Spots scattered
F. Additional oblique rows present
1. Above midlateral level +
2. Below midlateral level +
G. A large midlateral patch of minute
spots at one third of body length +
H. Scattered minute spots above and
below midlateral level +
Head pigment
I. small spots on throat
J. Small spots elsewhere on head
Gut pigment
K. Small spots ventrally before stomach
and dorsally behind stomach + + +
L. Small spots ventrally behind stomach (+) +
M. Series from throat to anus
1. Approx. one spot every 1-2 segments
2. Spots widely spaced (in young only) +
3. 6-9 groups of spots
Other pigment
N. Small spots on anal and dorsal bases + + +
O. A series of spots before dorsal fin;
few, large (young); many, small
(full grown) + +
Morphological:
P. Posterior teeth bladelike
Q. An "exterilium" intestine ( + ) + +
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
81
CONG Rl DAE
El
7^
z;!
MURAENESOCIDAE
.<;«t<g.ia;.T-^.7rrrfrnrf^i'^-?yi'^
K
M
T
Fig. 37. Illustrations accompanying Table 17.
82 ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 8. Pigment and Morphological Characters of Congridae (Congrinae), Colocongridae and Heterenchelyidae. + = All or most
species; ( + ) = some species only. Refer to Fig. 38.
Characters mycler congrus ger Conger
Bathy-
Aero-
Bathv
urocon
myaer
congrus
ger
Pseudo-
Pseiido-
Pan-
Gnalho-
Hilde-
xenomv-
phi-
Scala-
I'ro-
Colo-
Pviho-
lunch
phis
brandia
siax
chfhys
nago
conger
conger
nwhthys
ihys
Lateral pigment
A. A scattered row of spots below
midlateral level (rarely) (+)
B. A row of spots above and below
midlateral level ( + ) ( + )
C. Many, small spots scattered
below midlateral level ( + )
D. Single, small spots all over
lateral surface ( + )
E. Spots scattered in groups over
lateral surface
Axial pigment
F.
On spinal cord
G.
A single midlateral row
1. Spots widely spaced
2. I spot every 1-2 segments
3. 1 spot every segment
4. 2 spots every segment
5. spots widely spaced.
dendritic
H.
An extra, scattered row below
I.
3 deep spots postanally
(+)
(+) +
(+) + (+)
(+) + {+) + + {+)
(+)
(+)
Head pigment
J. A crescentic patch below eye +
K. Spots on throat +
Gut pigment
L. Spots in two regular rows
from throat to anus
1 . Close together (every
1-2 segments) + + + + +
2. Widely spaced (at least
in young)
M. Spots m clumps on gut loops +
Other pigment
N. Above anal base + ( + )
O. Along anal base + + + +
P. Along dorsal base + +
Morphological
Q. Some lower teeth bicuspid
R. Upper teeth needle-like
S. Lower teeth needle-like
-1-
+
(+)
-1-
-1-
+
-1-
-1-
+
-1-
+
+
-1-
+
+
+
+
+
+
+
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
83
Fig. 38. Illustrations accompanying Table 18.
84
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 19. Pigment and Myomere Characters of Nemichthyidae and Nettastomatidae.
* = larva unidentified. Refer to Fig. 39.
+ = All or most species; ( + ) some species only;
Characters
Avocel- Lahich-
ttna ihys*
Nemu'h-
thys
Facci-
olella
Neria- Netten- ?Netten-
sloma chelys chetys
Hop- Sauren-
lunnis chelys
I'ene-
Axial pigment
A. Deep on vertebral column
1. Single spot, or bipartite
2. Several spots along body
B. Small spots on top of spinal
cord, at least posteriorly
Head pigment
C. On snout and lower jaw
D. Deep behind eye
Gut pigment
E. A ventral row of minute
spots before stomach
F. A row of minute spots above
gut along its length
G. A patch of minute spots on
liver
H. A patch of minute spots
below kidney
I. Spots scattered between
liver and kidney patches
J. A regular longitudinal
series
(+)
(+)
+
+
+
+
+
+
(+)
(+)
+
+
+
+
+
+
+
+
(+)
(+)
Other pigment
K. Several groups of internal
spots along body subaxially
1. One or two in each group
+
2. Each group multiple (4)
+
L. Spots on ventral body wall
postanally
(+)
(+)
M. Minute spots on dorsal and
anal bases
+
+
+
Myomeres/vertebrae
Min.
177
174
ca.
238
186
209
ca.
192
ca.
Max.
216
191
400 +
294
246
273
257
276
224
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
85
N ETTAS TO MA TIDA E
NEMICHTHYIDAE
A2
Fig. 39. Illustrations accompanying Table 19.
86
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 20. Pigment, Morphological and Myomere Characters of Anguillidae, Derichthyidae and Serrivomeridae. + = All or most
species; ( + ) = some species only. Refer to Fig. 40.
Taxa
Characters
Anguilla Dertchlhys Nessorhamphiis Platuromdes Sernvomer Slemontdium
Lateral pigmeitt
A. Absent
B. Minute compact spots just below
midline on nearly every segment
C. Midline spots restricted to postanal
region (a few spots further forwards)
D. A series of minute spots on body wall
postanally
E. Minute spots on anal and dorsal bases
Head pigment
F. Absent
G. A cluster of minute spots in orbit
above eye
Morphological
H. Gut length
1. 0.7 total length
2. 0.75 total length
3. 0.9 total length
I. Dorsal fin origin
1. Just anterior to anus
2. Just behind midlength
3. At about midlength
J. Position of last vertical vessel
1. Behind mid-gut
2. Before mid-gut
Myomeres/vertebrae
Min.
Max.
(+)
+
(+)
(+)
-1-
100
119
126
134
135
139
153
170
147
169
137
141
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
Fig. 40. Illustrations accompanying Table 20.
88 ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 21. Pigment, Morphological and Myomere Characters of Muraenidae. + = All or most species; ( + ) = some species only;
unidentified. Refer to Fig. 41.
Taxa
Characters
Anarchias
Channo-
muraena
Urn-
prengius
Enchely-
core
Gyninii-
thorax
Muraena
Thvrso-
idea
Unidcn-
Ulied'
Uniden-
iihed*
Lateral pigment
A. Minute spots scattered over body
surface
+
Axial pigment
B. Small compact spots ventrally
on spinal cord, at least
posteriorly
Head
pigment
C.
Few to many, small, scattered spots
often compact
D.
Similar spots on brain
Gut pigment
E.
Ventral row behind stomach only
F.
Ventral row along whole of length
G.
Ventral row before stomach, dorsal
row behind stomach
H.
Short row before anus dorsally
1.
In disjunct groups ventrally
Other pigment
J.
Before dorsal base
K.
Along dorsal base
L.
Before anal base
M.
Along anal base
N.
Scattered over ventral surface
anteriorly
Morphological
O.
Dorsal and anal fins restricted
to tip of caudal
P.
Dorsal origin at myomere 30-40
Q.
Dorsal origin at myomere 40-75
+ +
+ +
(+) (+)
+
(+)
+ + +
+ + +(+)+ +
(+) +
+ + +(+)+ +
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
89
Fig. 4 1 . Illustrations accompanying Table 2 1 .
90
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 22. Pigment and Myomere Characters of Xenocongridae. + = All or most species; (+) some species only;
Refer to Fig. 42.
larva unidentified.
Cates- Chtlo-
bya rhinus
CMop-
Kau- Powell-
pichthys ichthys*
Robin-
Xeno-
conger*
Uniden-
tified*
I'niden- Uniden- Thalassen-
tifted* lifted* chdys
Lateral pigment
A. Absent (first pair maxillary and
mandibular teeth very large)
Midlateral pigment present
B. Irregular double row of minute
spots along body
C. One minute spot per segment
D. Round groups of minute spots
along body
E. Large spots, widely spaced
F. Axial spots confined posteriorly
Pigment elsewhere
G. W-shaped rows of minute spots on
anterior margin of segments
H. Round groups of minute spots all
over body
Head pigment
I.
(+)
Scattered spots behind eye and on
heart
A row of spots along upper and lower
jaws
A patch below iris
A few spots on snout tip or on
olfactory organ
Gut pigment
M. Minute spots ventrally before
stomach and dorsally behind stomach
Minute scattered spots dorsally along
gut to anus or only postenorly
Minute spots below gut
Large widely spaced spots
Minute spots on liver
Round groups of spots on gut
Other pigment
S. Spots on anal base or rays
Myomeres/vertebrae
Min.
Max.
J.
K.
L.
N.
O.
P.
Q.
R.
+
(+) +
+ +
+ +
136 98 116 97
141 107 5 125
130
136
ca.
157
142
163
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
91
Fig, 42. Illustrations accompanying Table 22.
92
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 23. Pigment and Myomere Characters of Moringuidae, Cyematidae and Saccopharyngoidea. +
tt = Castle and Raju, 1975; * = Larva unidentified. Refer to Fig. 43.
All species; t = Smith, 1979;
Taxa
Monn-
Neo-
Uniden-
Neo- L holli
Sacco-
Eury-
Mono-
Uniden-
Characters
H"a
conger
tifiedt
Cyema
cyema* group*
pharynx
pharynx
gnalhus
lifiedtt
Lateral pigment
A. Absent
+
+
+
+
+
B. Scattered over lateral surface
+
C. Midlateral series
1. Single spot behind anus
+
2. Multiple spots along body (5-1 1)
+
+
D. Short rows of spots on myotomes
dorsally and ventrally (juvenile
pigment)
+
Head pigment
E. On snout and lower jaw
1. Present
+
+
2. Absent
+
+
+
+
+
+
+
Gut pigment
F. One large posterior spot
+
+
+
G. One small anterior spot
+
H. Series of spots along gut
+
+
I. Minute spots scattered over
posterior surface
+
+
+
+
J. Few small spots on liver
+
Myomeres/vertebrae
Min.
107
96
121
74
102
138
97
94
ca.
Max.
180
110
132
78
104
250
125
113
105
CASTLE: NOTACANTHIFORMES, ANGUILLIFORMES
93
Fig. 43. Illustrations accompanying Table 23.
Elopiformes, Notacanthifomies and Anguilliformes: Relationships
D. G. Smith
NOTACANTHIFORMES
THE Notacanthiformes is composed of two clearly defined
families, the Halosauridae and Notacanthidae. Overall, the
Halosauridae is the more primitive family. McDowell (1973)
divided it into two subfamilies: the Halosaurinae, containing
only Halosaurus, and the Halosauropsinae, containing Halo-
sauropsis and Aldrovandia. The notacanthids show a number
of specializations not found in the halosaurs, involving mainly
the mouth and dorsal fin. The Notacanthidae contains either
two or three genera, depending on the placement of Lipogenys.
McDowell recognized only Nolacanlhus and Polyacanthonolus
in the Notacanthidae while assigning Lipogenys to a separate
family. He considered the Lipogenyidae and Notacanthidae to
form a suborder of the Notacanthiformes, the Notacanthoidei,
which stood opposed to the Halosauroidei. Greenwood (1977),
however, felt that Lipogenys was closely related cladistically to
Polyacanthonolus and that those two genera formed the sister
group of Notacanlhus. A classification of the Notacanthiformes
based on Greenwood's interpretation would be as in Fig. 44.
Notacanthiform larvae cannot yet be identified confidently
below the ordinal level and hence can tell us little about rela-
tionships within the order. Smith (1970) gave reasons to suspect
that the Tiluropsis form (short head, vertically elongate eye)
belongs to the Halosauridae. Circumstantial evidence suggests
that the Tiluriis form (short head, normal eye) is the larva of
the Notacanthidae. Tilunis is the only notacanthiform larva
found in the Mediterranean. Although adult notacanthids of
both Notacanlhus and Polyacanthonolus occur in the Mediter-
ranean, halosaurs apparently do not (McDowell, 1973). The
identity of the third basic type of notacanthiform larva, known
as Leptocephalus giganteus (long head, normal eye), cannot even
be guessed at this point.
Anguilliformes
The Anguilliformes, the true eels, is the largest and most
specialized of the elopomorph orders. A definitive classification
of the Anguilliformes does not yet exist. The scheme that follows
can be considered an outline that will be filled in and modified
as studies continue.
The eels can be divided into two groups: those in which the
frontal bones are fused, and those in which they remain as
separate right and left elements. This observation dates back to
Regan (1912), but its phylogenetic significance has not always
been agreed upon. Regan himself said nothing about it one way
or another; he simply used it as a key character. A case can be
made for the view that the fusion of the frontals was a single
event that occurred quite early in the evolutionary history of
eels and therefore reflects a real phylogenetic division. On the
whole the fused-frontal group contains more primitive members
than the divided-frontal group, although the fused condition is
itself a derived character state. Except for Anguilla. all the di-
vided-frontal eels are markedly specialized, including pelagic
and fossorial representatives. Yet in none of these lines has a
fusion of the frontals been among the modifications. Of the more
specialized members of the fused-frontal group, all but the Ser-
rivomeridae can be clearly traced back to more primitive mem-
bers, all with perfectly fused frontals. It is more parsimonious
to assume that fusion took place once at a point early in an-
guilliform evolution than to assume that it occurred several
times early but not at all later on.
The number of families in the fused-frontal group is still
somewhat uncertain. Ten are provisionally recognized here. The
Synaphobranchidae, Simenchelyidae, and Dysommatidae are
closely related and could easily be considered subfamilies of the
Synaphobranchidae (Robins and Robins, 1976). They combine
some very primitive characters with some peculiar specializa-
tions and do not seem to be intimately related to any of the
other families. The Nettastomatidae shares several advanced
characters with certain congrids and could be considered a de-
rivative of that group. The interrelationships of the remaining
families are not clear; the resemblances involve mainly primi-
tive characters. The Ophichthidae is a large and morphologically
diverse family containing both generalized and highly modified
forms. It is united by certain specialized characters such as a
ventrally displaced posterior nostril, a reduced caudal fin, and
numerous branchiostegal rays that overlap on the ventral mid-
line. The Congridae (including Macrocephenchelyidae) is also
a large family, but without the extreme variety of external mor-
phology found in the Ophichthidae. Its specializations are more
subtle and consist mainly of trends in several characters. The
Colocongridae and Muraenesocidae have at various times been
included in the Congridae, but again the resemblances are main-
ly in primitive characters. Neither family fits the pattern of
character modification found in the Congridae, and both show
at least one primitive character that is absent in nearly all con-
grids: separate hypohyals. The Muraenesocidae is here restricted
to Muraenesox itself and its close relatives Congresox, Cyno-
ponticus and Sauromuraenesox. Of the other genera previously
referred to this family, Hoplimnis has been removed to the
Nettastomatidae (Smith, 1979; Smith and Castle, 1982), and
Xenomystax (including Paraxenomyslax) probably belongs in
the Congridae. The Derichthyidae and Serrivomeridae are mid-
water eels, the former relatively little modified, the latter highly
modified. The Serrivomeridae was formerly associated with the
Nemichthyidae, but this seems unlikely. The completely fused
frontals and massive palatopterygoid arcade of serrivomerids
difier strikingly from the partially fused frontals and reduced
pterygoid found in nemichthyids.
There are eleven families of eels with divided frontals: the
Anguillidae, Moringuidae, Heterenchelyidae, Myrocongridae,
Xenocongridae, Muraenidae, Nemichthyidae, Cyematidae, Sac-
copharyngidae, Eurypharyngidae, and Monognathidae (the
monognathids actually have fused frontals, but they are clearly
related to the saccopharyngids and eurypharyngids and the fu-
sion seems secondary). Although they are more clearly defined
than the fused-frontal families, their interrelationships are still
uncertain. Except for the Anguillidae, they are all distinctly
specialized, either for burrowing (Moringuidae, Heterenchelyi-
dae), for midwater life (Nemichthyidae, Cyematidae, Sacco-
94
SMITH: ELOPIFORMES, NOTACANTHIFORMES AND ANGUILLIFORMES
95
HALO-
5AURU5
HAL05AUR-
0P5I5
ALDRO-
VANDIA
POLY AC AN -
TH0N0TU5
LIPO-
0ENY5
NOTA-
C A NTH US
HAL05AURINAE
NOTACANTHINAE
HAL05AUR0P5INAE
HALOSAURIDAE
NOTACANTHIDAE
Fig. 44. Hypothesis of relationships within the Notacanthiformes.
Fig. 45. Leptocephali of Neoconger (above) and Moringua (below) (Moringuidae).
96
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 46. Heads of leptocephali of Dysommatidae (above) and Syn-
aphobranchidae (below), showing telescopic eye.
pharyngidae, Eurypharyngidae, Monognathidae), or as cryptic
forms with modified lateral-line and gill-arch characters (My-
rocongridae, Xenocongridae, Muraenidae). Two clear associa-
tions are evident within this group. One contains the Myrocon-
gridae, Xenocongridae, and Muraenidae. These three families
are relatively generalized externally but share a marked reduc-
tion in gill-arch elements and in the lateral line. The second
association contains the three families Saccopharyngidae, Eu-
rypharyngidae, and Monognathidae, the so-called gulper eels.
These are highly modified midwater eels with a greatly enlarged
mouth and an elongated, posteriorly directed suspensorium.
The gulpers show extreme reduction in all the skeletal elements,
and their relationship to other eels is difficult to determine.
Among the remaining families, the Anguillidae is quite primi-
tive morphologically, but it seems to have no advanced char-
acters clearly linking it to any of the other families. The Mo-
ringuidae and Heterenchelyidae are fossorial forms that never-
theless show substantial internal differences from each other
(Smith and Castle, 1972). Their resemblances may simply be
convergent adaptations to a similar way of life. The Nemichthy-
idae and Cyematidae both have prolonged, nonocclusible jaws
studded with liny recurved teeth, but they differ markedly in
almost every other character; their traditional association must
be questioned.
Larval characters have so far proved more useful in eluci-
dating relationships within families than between them. Some
examples will illustrate the contribution that larvae have made
to systematics.
The Moringuidae consists of two genera, Moringna and Neo-
congcr. Although both are basically fossorial forms, they differ
enough in external appearance that for more than a century they
were placed in different families. It was only the striking simi-
larity of the larvae (Fig. 45) that prompted a critical comparison
of the adults (Smith and Castle, 1972). In this case, the larvae
show the relationship much more clearly than do the adults.
The close relationship between the Synaphobranchidae and
Dysommatidae is supported by a unique feature of the larvae—
the telescopic eye (Fig. 46).
The genus Hoplimnis has long been placed in the family Mu-
raenesocidae because of its possession of a pectoral fin and its
enlarged median vomerine teeth. Saiirenchelys was always con-
sidered a nettastomatid because it lacked a pectoral fin. Smith
and Castle (1982) showed that the larvae of these genera are
indistinguishable (Fig. 47). On that basis and because of many
similarities in the adults, Hophinnis and Saurenchelys were shown
to be closely related and to belong in the Nettastomatidae. The
two characteristic swellings in the gut of larval Hoplunnis and
Saurenchelys are also found in the larvae of Nettastoma and
Nettenchelys.
The major problem in eel systematics today is the relationship
between the families, and here larvae provide little help. Sim-
ilarities occur between larvae of families which otherwise show
no evidence of close relationship. For example, the larvae of
the Anguillidae and Derichthyidae are quite similar (the larva
of Derichthys was even named Lcplocephalus angiulloides). but
the two families do not seem especially close and fall on opposite
sides of the fused-frontals vs. divided-frontals dichotomy. The
larvae of the Heterenchelyidae resemble those of certain con-
grids, but heterenchelyids have divided frontals and congrids
have fused frontals. Larvae of the congrid genus Acromycter
Fig. 47. Leptocephali of Hoplunnis tenuis (above) and Saurenchelys sp. (below) (Nettastomatidae).
SMITH: ELOPIFORMES. NOTACANTHIFORMES AND ANGUILLIFORMES
97
Fig. 48. Leptocephali of Cyema alrum (Cyemalidae) (above) and Nemichthys scolopaceus (Nemichthyidae) (below).
98
ONTOGENY AND SYSTEMATICS OF HSHES- AHLSTROM SYMPOSIUM
ELOPIDAE
MEGAL -
OP I DAE
ALBUL-
IDAE
HALO-
SAURIDAE
NOTA-
CANTHIDAE
EELS-
21 FAM5.
ELOPIFORMES
ANGUILLOIDEI
Fig. 49. Hypothesis of relationships between major groups of elopomorphs.
(Fig. 52E) have a looped gut and superficially resemble certain
ophichthids (Fig. 51 A); on the other hand, some ophichthid
larvae (for example, Basicanichthys. Fig. 52D) have a weakly
looped gut and superficially resemble congrid larvae (Smith and
Leiby, 1980).
A contraindication of relationship may be shown by the larvae
of the Nemichthyidae and Cyematidae. It was mentioned above
that these two families differ in many characters and that their
traditional association must be questioned. The larvae of these
families are as different from each other as any two leptocephali
can be. Nemichthyid larvae are long and slender with a simple
gut that reaches almost to the tip of the tail. Cyematid larvae,
on the other hand, are high and deep and their gut contains
several characteristic loops (Fig. 48). Some observers have no-
ticed a resemblance between cyematid larvae and saccopha-
ryngoid larvae and have suggested that these families are related
(Benin, 1937; Raju, 1974).
Despite the caveats that must be invoked when dealing with
the systematic implications of leptocephali, these larvae play an
important role in systematic studies of eels. They provide ad-
ditional characters to be used in systematic analysis, and they
are often more readily accessible than adults. The cryptic or
burrowing habits of most adult eels make them difficult to collect
in large numbers. The larvae, on the other hand, live in open
water near the surface and can easily be collected with plankton
nets or midwater trawls. In many cases, larvae provide data on
distribution and species structure that are unavailable from adults
(Smith and Castle, 1972, 1982).
Elopomorphs
The Notacanthiformes and Anguilliformes belong to a group
of fishes called elopomorphs, along with the Megalopidae, Elo-
pidae, and Albulidae (including Pterothrissidae). Current con-
cepts of the interrelationships of the major groups of elopo-
morphs are illustrated in Fig. 49 (Greenwood, 1977; Patterson
and Rosen, 1977; Lauder and Liem, 1983). The trichotomy
exists because there seem to be no derived characters that clearly
link any of the three main branches with any of the others.
Elops and Megalops (including Tarpon) seem more similar
to each other than either is to Alhida. but this may be because
they are both midwater feeders with terminal mouths, whereas
A/hula is a bottom feeder. Alhula has several specializations
(enlarged cephalic canals, prolonged snout) that are lacking in
Elops and Megalops. Most if not all of the resemblances between
Elops and Megalops may be explained either as primitive char-
acters or as adaptations to a similar way of life. Megalops has
several derived characters not found in Elops, most notably the
vascular air bladder and the otophysic connection. Elops does
not seem to have any feature that is derived relative to other
elopomorphs.
Several synapomorphies can be cited to link the Notacanthi-
formes and the Albulidae (Nelson, 1973; Greenwood, 1977).
The eels are usually placed on the albulid branch as well, but
this is still an open question. The Anguilliformes and Notacan-
thiformes share a similar elongate body form, but this feature
has evolved so many times in fishes that it means little by itself
SMITH: ELOPIFORMES, NOTACANTHIFORMES AND ANGUILLIFORMES
99
Fig. 50. Caudal structure of an anguilliform leptocephalus (above) and a notacanthiform leptocephalus (below).
The only real character seems to be the swim-bladder mor-
phology of the two groups (Marshall, 1962). but a critical com-
parison with the swim bladders of Elops and Megalops has not
been made. Until that is done, it cannot be determined whether
the swim bladders of eels and notacanthiforms represent a syn-
apomorphy or simply a general condition of elopomorphs.
Larvae probably cannot resolve the trichotomy. A classifi-
cation based on larvae would also yield three groups, but they
would not be the same three groups. The three main groups of
larvae are the fork-tailed group, the notacanthiform group, and
the anguilliform group. These simply represent the condition in
the adults. The forked tail is a primitive condition retained in
the Elopidae. Megalopidae. and Albulidae.
Larvae do not reveal much about relationships within the
fork-tailed group either. The larvae of Elops and Megalops re-
semble each other more than they do that of Albula. They are
smaller, the gut is shorter, and the dorsal fin is above or nearly
above the anal fin. Albula shows a trend toward elongation,
although the myomeres are no more numerous than those of
Elops. The gut is very long, ending under the hypural, and the
dorsal fin is much farther forward than the anal fin. Pterothrissus
is even more elongated and grows larger before metamorphosis
than Albula. In albulids the myomeres are more V-shaped than
W-shaped. If the primitive condition is small size and relatively
short larval life, then Megalops has the most primitive larva. It
is the smallest known leptocephalus, metamorphosing before it
reaches 30 mm standard length, at an age of two to three months
(Smith, 1980). Larvae of Elops are closer in size and form to
those oi Megalops than to Albula, but this does not necessarily
demonstrate that the two former genera are more closely related
to each other cladistically than either is to Albula. It could simply
mean that Elops and Megalops retain a more primitive larval
form and that, once again, they merely lack a specialization
found in albulids.
The larvae of the Notacanthiformes and Anguilliformes do
not indicate a particularly close relationship between the two
groups. The elongated form simply reflects the condition in the
adults, and in several respects the two groups are quite different.
The short-based dorsal fin and the presence of pelvic fins in
notacanthiform larvae immediately separate them from an-
guilliform larvae. Eels lack pelvic fins and their dorsal fin is long
and confluent with the caudal and anal fins. In both these char-
acters the notacanthiforms show the more primitive state. In
the structure of the tail, however, the notacanthiforms are more
highly modified. Eels, despite their elongate form, retain a caudal
fin complete with hypural plates and caudal fin rays. To be sure,
the caudal fin is greatly reduced and shows much fusion of
elements, but it clearly exists, in larvae as well as adults (Fig.
100
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 5 1 . Leptocephali of (A) Callechelys sp.; (B) Catesbya pseudomuraena, and (C) Kaupichlhys hypoproproides.
50, top). Notacanthiforms have no true caudal fin. In adult
notacanthiforms the vertebrae approaching the tip of the tail
become progressively less ossified, the centra being reduced to
rings around the notochord separated from the neural and hemal
arches. Finally the vertebrae disappear, leaving the notochord
freely exposed (McDowell, 1973). There is no hypural structure,
and caudal fin rays, if they exist, are indistinguishable from the
posterior anal fin rays. The notacanthiform larva likewise has
no caudal fin (Fig. 50, bottom); the notochord ends freely, but
there are two structures that may be hypural elements. Posterior
to these and to the notochord is a single filament that trails
freely for a variable distance and might represent a caudal fin
ray. The anal fin occupies the short space between the anus and
the end of the tail proper (excluding the caudal filament). The
important point here is that lumping notacanthiform and an-
guilliform larvae as pointed-tail leptocephali is unwarranted,
because the caudal structure is quite different in the two groups.
Returning to the diagram in Fig. 49, the fork-tailed leptocephali
can be viewed as the primitive type of leptocephalus present in
the elopid and megalopid branches and retained in the Albulidae
as well. Two pomts of transformation occur, one in the nota-
canthiform line and one in the anguilliform line. The modifi-
cations in each reflect modifications in the adults and by them-
selves are not indications of a special relationship. Additional
leptocephali illustrations were prepared and are presented here
without further comment (Figs. 51, 52).
Relationships between Elopomorphs
AND OTHER TeLEOSTS
A widely favored view today is that the teleosts consist of
four major groups in a cladistic sense: the Osteoglossomorpha,
Elopomorpha, Clupeomorpha, and Euteleostei (Greenwood et
al., 1966; Greenwood, 1973; Nelson, 1973; Patterson and Ro-
sen, 1977). These groups are arranged in a hierarchy with the
Osteoglossomorpha as the sister group of the remaining three,
the Elopomorpha as the sister group of the remaining two, and
the Clupeomorpha as the sister group of the Euteleostei (Fig.
53). This classification is based on a few characters that are
thought to represent synapomorphies. It is essential, therefore,
to evaluate these characters carefully, because the whole clas-
sification stands or falls on their reliability.
The Elopomorpha is united by three characters: I ) the pres-
ence of rostral and prenasal ossicles; 2) the initial fusion of the
angular and retroarticular bones in the lower jaw; 3) the presence
of a leptocephalus larva. It is not certain that eels have rostral
ossicles. Considering the extreme fusion that has taken place in
the anterior extremity of the skull in eels, it should not be
surprising if the rostral ossicles were lost as well. Still, it means
that the character may not be wholly inclusive of the group. The
second character, the fusion of the angular and retroarticular,
seems to hold for eels (Leiby, 1979b) and appears to be a true
synapomorphy. That leaves the leptocephalus, and its role is
SMITH: ELOPIFORMES. NOTACANTHIFORMES AND ANGUILLIFORMES
Fig. 52. Leptocephali of (D) Bascamchthys sp.; (E) Acromycter sp.; (F) Hildebrandia; and (G) Dysomma anguillare.
crucial. If it is a synapomorphy, then the congruence between
it and the lower-jaw character reinforces the naturalness of the
Elopomorpha. Furthermore, it is a more complex character, thus
less likely to show parallelism than a simple process like the
fusion of two bones in the lower jaw (which, indeed, has hap-
pened independently in some osteoglossomorphs).
To explore this matter, we must first establish clearly what a
leptocephalus is. If, as some have maintained, it were simply a
ribbon-like larva with a posterior anus and a dorsal fin that
moves forward at metamorphosis, then it would tell us little
about elopomorph phylogeny. Many lower teleosts have such
larvae. A leptocephalus is considerably more than this, however.
102
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
0STE0GL05 -
SOMORPHA
ELOPO-
MORPHA
CLUPEO-
MORPHA
EUTEL -
EOSTEI
Fig. 53. Hypothesis of relationships between major groups of Tele-
ostei.
The unique structure of a leptocephalus can be appreciated best
in cross section (Fig. 54, left). The viscera lie along the ventral
margin in a narrow strand. The notochord, dorsal nerve cord,
and dorsal aorta lie together in the longitudinal axis of the body
about midway between the dorsal and ventral margins. The
myomeres form a thin layer on the outside. Filling the rest of
the interior of the body is an acellular mucinous material bound-
ed by a continuous layer of epithelial cells. The mucinous pouch
separates the viscera, the notochord and the two sides of the
body musculature from each other and gives form and rigidity
to the body. The characteristic shrinkage of the leptocephalus
at metamorphosis is due to the loss (presumably by resorption)
of the internal mucinous material. A typical clupeid larva such
as Elnaneus teres (Fig. 54, right) is constructed much differently.
Here there is no mucinous pouch. The notochord occupies a
large part of the cross-sectional area and is surrounded imme-
diately by the thick axial musculature to form a solid, compact
structure. The viscera lie immediately below the dorsal aorta.
Leptocephali have a small head and a set of long, sharp teeth
whose function is uncertain, since leptocephali do not seem to
be predatory. The basic structure of a leptocephalus is the same
whether it is an elopiform, notacanthiform or anguilliform. A
leptocephalus larva is known for every family of elopomorphs
except the rare, monotypic Myrocongridae, so the character
seems entirely inclusive of the group. Nothing even remotely
comparable is found outside the Elopomorpha.
Fig. 54. Cross section through the bodies of a leptocephalus (Meg-
alops allanlictis) (left) and a clupeid larva (Elrumeus teres) (right). DA,
dorsal aorta; NC, notochord; SC, spinal cord.
The leptocephalus, then, must be considered a true synapo-
morphy and powerful evidence in favor of the monophyly of
the Elopomorpha. Perhaps nowhere else in fish systemalics have
larval stages played a more important role.
The Marine Biomedical Institute, The University of Texas
Medical Branch at Galveston, 200 University Boule-
vard, Galveston, Texas 77550.
Ophichthidae: Development and Relationships
M. M. Leiby
THE family Ophichthidae, comprising approximately 250
nominal species and 53 recognized genera, is arranged in
six tribes and two subfamilies (McCosker. 1977) (Fig. 55). The
subfamilies, Myrophinae and Ophichthinae, are separated by a
number of characters. All adult Myrophinae have a well-de-
veloped caudal fin which is continuous with the dorsal and anal
fin. Adult Ophichthinae, except for Echelus in the tribe Ophich-
thini and Lcptenchelys in the tribe Bascanichthyini, lack a caudal
LEIBY: OPHICHTHIDAE
103
Family Ophichthidoe
Subfomily Myrophmoe
Sub fomily Ophichthmae
I— II
Ttib«
Myrophin
Tribe
Ophichthini
Tribe
Sphogebranchir
Tribe
Basconichthyini
Tribe
Callechelyini
£
>■
C
£
o
■6
3
w
£
3
m
■^
t
o
:^
o
e
V
b
■a
3
c
o
c
at
CT
i
3
tl
W
3
3
c
i
3
E
a
o
o
E
o
o
o
«
w
•
3
u
a.
u
a>
(D
Z
3
<3
s
VI
o
tn
O
O
a.
u
UJ
^
-*
o = •)
w ^ tt
-o ^ •» ly
Callechelyin
Ancestor
Bosconichf hy s-like
Ancestor
Moderately Specolized
Optiictittiin-like Ancestor
oderotely Speciolized
Ophictttttin-like Ancestor
Moltfolioptiis Or
E vips- like Ancestor
Ouossiremus-like Ancestor
Ance strol Myroph
Ancestrol Optlictlttlin
= Tribe BenThenchelyin
Congrid-like Ancestor
Fig. 55. Hypothesized relationships of the subfamilies and genera of the eel family Ophichthidae.
tin having instead a hardened tail tip with, at most, a few ru-
dimentary caudal rays embedded in the flesh of the tail. The
monotypic genus Leptenche/ys. known only from the 1 1 5 mm
type specimen, has caudal-fin rays, but they are weakly devel-
oped compared to those of a myrophin (McCosker, 1977). Since
all ophichthid larvae have a well-developed caudal fin until the
onset of metamorphosis, the presence of weakly developed rays
in the only known specimen of Leplenchelys may be an anomaly
resulting from incomplete resorption during metamorphosis.
The well developed caudal fin of Echelus has prompted most
earlier authors to place it in the family Echelidae (=Ophichthi-
dae, in part) or to ally it with the subfamily Myrophinae (e.g..
Dean, 1972; Blache, 1977); however, the osteology of the genus
(McCosker, 1977) and its larval morphology (Blache, 1977: Figs.
72 and 74) clearly place Echelus in the subfamily Ophichthinae
and ally it with the tribe Ophichthini.
Adult Myrophinae have four to seven branchiostegal rays
attached to the epihyal and ceratohyal and 1 3-45 free (unat-
tached) branchiostegal rays which originate posterior to the tips
of the epihyals. Most adult Ophichthinae have the majority of
their branchiostegal rays attached to the epihyal and ceratohyal.
The free branchiostegal rays of all Ophichthinae originate an-
terior to the tips of the epihyals.
The ceratohyal, epihyal and hypohyal of both the Myrophinae
and the Ophichthinae originate from a single block of cartilage
with the first center of ossification being a thin strip along the
lateral face of the cartilage (Leiby, 1979a, b; 1981). When de-
velopment is complete, the ceratohyal of the Myrophinae is a
simple bone which terminates about midpoint along the lateral
face ofthe epihyal (Dean, 1972; McCosker, 1977; Leiby, 1979b).
The ceratohyal ofthe Ophichthinae has a slender, elongate distal
portion which terminates about midpoint along the lateral face
of the epihyal and a medial portion which is attached to the
proximal end ofthe epihyal by a cartilage (McCosker, 1977;
Leiby, 1981).
The urohyal ofthe Myrophinae and Ophichthinae ossifies in
a bifurcated medial ligament which is attached to the developing
hypohyals. In the Myrophinae, the urohyal is generally limited
104
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
L L-
Fig. 56. (Upper.) Anterior portion of Myrophis punclalus larva depicting typical myrophin gut morphology. Abbreviations: LL|_j, liver lobes
1-3; GB, gall bladder. (Lower.) Anterior portion of Neenchelvs microlrelus larva depicting gut morphology. Abbreviations: LL,.,, liver lobes 1-
2; GB. gall bladder.
to a basal plate which ossifies from the hypohyal to the bifur-
cation of the ligament. The urohyal of the Ophichthinae gen-
erally ossifies to include a spike which extends well posterior to
the area of the bifurcation.
The gill openings of the Myrophinae are midlateral and con-
stricted. Ophichthine gill openings are variable in position, their
major axis ranging from midlateral to ventral, but always un-
constricted.
Leptocephali belonging to five of the nine myrophin genera
have been identified. Larvae of four of these five genera have
three unconnected liver lobes with the gall bladder on the third
lobe (Fig. 56-upper). Larvae of the fifth genus, Neenchelvs. which
differ trenchantly from all other ophichthid larvae, have two
unconnected liver lobes with the gall bladder on the second lobe
(Fig. 56-lower). Leptocephali belonging to twenty of the forty-
four ophichthin genera have been identified. All twenty of these
genera have two connected liver lobes with the gall bladder on
the second lobe (Fig. 57-upper).
LEIBY: OPHICHTHIDAE
105
5mm
Fig. 57. (Upper.) Anterior portion of Ophichthus gomesi larva depicting typical ophichthin gut morphology. Abbreviations: LL|_2. liver lobes
1-2; GB, gall bladder. (Lower.) Middle portion of Ophichthus gomesi larva depicting position of nephros relative to anus in some members of
the Ophichthus lineage of the tribe Ophichthini. Abbreviations: N, nephros; A, anus.
The dorsal fin of known myrophin lai^ae has well-developed
pterygiophores and fin rays prior to the onset of metamorphosis
and migrates only a few myomeres anteriorly (4-6) during meta-
morphosis to reach its adult position. The dorsal fin of known
ophichthin larvae, which is weakly developed having only pte-
rygiophores and rudimentary rays in its anterior portion prior
to metamorphosis, must migrate 5-20 myomeres anteriorly dur-
ing metamorphosis in species having the dorsal fin antenor to
the branchial aperture as adults, and 20-50 myomeres in species
having the dorsal fin posterior to the branchial aperture as adults,
and is resorbed m species which are finless as adults.
The subfamily Myrophinae contains two tribes (sensu
McCosker, 1977), the Myrophini and the Benthenchelyini. Os-
teological examination of adults in the tribe Myrophini indi-
cated the presence of three lineages consisting of Pseudomyro-
phis and Neenchelys; Myrophis, Ahlia. and a currently
undescribed genus; and Muraemchlhys and its allies. The My-
rophis and Muraemchlhys lineages share a common ancestor
(Fig. 55). Larval morphology oi Myrophis, Ahlia and Muraen-
ichthys is very similar and supports the determination of a close
relationship for the two lineages. Larvae of these three genera
have three unconnected liver lobes, similar gut and opistho-
nephros morphology, and similar body length to depth ratios
(Fahay and Obenchain, 1978; Leiby, 1979b; Ochiai and No-
zawa, 1980). Pseudomyrophis larvae have three unconnected
liver lobes and a body length to depth ratio which is similar to
that of the Myrophis and Miiraenichthys lineages, but gut and
opisthonephros morphology is significantly different from that
seen in the Myrophis and Muraemchthys lineages and supports
the conclusion drawn from adult data that the Pseudomyrophis
lineage is distinct from the Myrophis and Muraenichthys lin-
eages. Nelson ( 1 966a) suggested that Pseudomyrophis micro-
pinna, the type of the genus, was congeneric with Neenchelys
hiutendijki, but that P. nimius, while belonging to the same
lineage, was separable at the generic level from either of the
other two species. Dean (1972) also felt that the differences
106
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
between P. micropinna and P. nimiiis warranted a separate ge-
nus for P. nimius. However, McCosker (1977, 1982) demon-
strated that Pseudomyrophis and Neenchelys are both valid gen-
era and that P. micropinna, P. nimius, P. atlanlicus and an
undescribed Pseudomyrophis from the eastern Pacific are con-
generic. Dean (1972) indicated that Myrophis frio properly be-
longs in the Pseudomyrophis lineage. Evidence from larval mor-
phology supports McCosker's (1977, 1982) recognition of
Pseudomyrophis and Neenchelys as valid genera, and supports
the recognition of P. micropinna, P. nimius, P. atlanlicus, the
undescribed Pseudomyrophis from the eastern Pacific, two un-
described Pseudomyrophis known only from their larvae in the
western Atlantic, one undescribed Pseudomyrophis from the
eastern Atlantic known only from its larva and erroneously
identified as P. nimius (Blache, 1977), and Myrophis frio as
congeneric. Pseudomyrophis larvae are readily distinguishable
from all other ophichthid larvae by a combination of the fol-
lowing characters: three unconnected liver lobes, undulating gut
and nephros, characteristic head shape, and pigmentation
(Blache, 1977; Leiby, in press a). Neenchelys larvae differ tren-
chantly from Pseudomyrophis larvae in having two, rather than
three, unconnected liver lobes, a gut lacking the marked un-
dulations seen in Pseudomyrophis larvae, and a much deeper
body than any other known ophichthid (Castle, 1 980; this paper.
Fig. 56-lower). Studies of adult Pseudomyrophis and Neenchelys
have clearly demonstrated that the two genera are more closely
related to each other than either is to any other genus ( McCosker,
1977, 1982). In the light of this information, the most parsi-
monious interpretation of the data on the larval morphology of
the two genera is that Neenchelys was derived from Pseudo-
myrophis or a Pseudomyrophis-hke ancestor. Pseudomyrophis
and all other known myrophin larvae except Neenchelys have
three unconnected liver lobes and similar body length to depth
ratios. It seems likely, therefore, that larvae of the ancestral
myrophin also had three unconnected liver lobes and a similar
body length to depth ratio. Neenchelys larval morphology can
be easily derived from this proposed ancestral larval morphol-
ogy by significantly deepening the body and foreshortening the
gut so that one liver lobe is lost. Derivation of Pseudomyrophis
larval morphology from a Neenchelys-Wkc ancestor requires a
change from the ancestral larval morphology body plan to the
Neenchelys larval body plan and a later re-emergence of the
ancestral larval myrophin body plan in Pseudomyrophis.
Benthenchelys cartieri. a highly specialized pelagic eel (Castle,
1972) is the sole member of the tribe Benthenchelyini. The
larvae of this species have not yet been described, but based on
the hypothesized evolutionary history of the Ophichthidae (Fig.
55), it seems likely that the larvae of 5. cartieri will have three
unconnected liver lobes, a well-developed dorsal fin which mi-
grates little during metamorphosis, and a body length to depth
ratio that is typical of the Ophichthidae. Discovery of these
larvae should help clarify relationships within the Myrophinae.
The subfamily Ophichthinae contains four tribes (sensu
McCosker, 1977); the Ophichthini, Sphagebranchini, Bascan-
ichthyini and Callechelyini. The tribe Ophichthini lies at the
evolutionary base of the subfamily Ophichthinae, and contains
the most primitive, least specialized members of the subfamily.
The ancestral ophichthin was probably Ophichthus-hke. The
tribe Ophichthini, which contains two lineages, and the tribe
Sphagebranchini can be easily derived from the generalized
ophichthin character states which are represented in the genus
Ophichihus (sensu McCosker, 1977). One lineage in the tribe
Ophichthini appears to be directly derived from the generalized
Ophichthus condition. The genus Echelus has been represented
as belonging to its own unique lineage in the Ophichthinae and
has been considered the most primitive member of the tribe
Ophichthini because in addition to having all the primitive
characters of its closest relative Ophichthus. it possesses a well-
developed caudal fin. A re-examination of adult Echelus char-
acters in conjunction with the larval characters oi Echelus sug-
gests, however, that Echelus belongs to the Ophichthus lineage
and that the caudal fin of Echelus is either a case of character
reversal or paedomorphosis which resulted in Echelus retaining
the larval caudal fin rather than losing it, as is apparently the
case in all other members of the Ophichthinae. In addition to
the generalized genera Echelus. Ophichthus, and Ophisurus, the
Ophichthus lineage contains two groups of specialized genera
which are closely tied to Ophichthus by a nearly continuous
character series. The Pisodonophis-Myrichthys-Cirrhimuraena
group differ from the basic Ophichthus body plan by having an
increased number of branchiostegals, multiserial dentition, and
individual sp)ecializations found in each genus. The second group,
containing Mystriophis and seven allied genera, are specialized
for the capture of large active prey by having a strengthened
suspensorium and enlarged dentition. The close relationship of
this group to Ophichthus is emphasized by similar adaptations
in some species of Ophichthus (McCosker, 1977). The close
relationship of the Ophichthus lineage is further emphasized by
the unique positioning of the nephros relative to the anus found
in many members of this lineage. Larvae from seven of the
fourteen genera in the Ophichthus lineage have been identified.
While there is considerable inter- and intrageneric variability
in the general morphology of these larvae, five of the seven
genera (Echelus. Ophichthus, Ophisurus, Echiophis, and Apla-
tophis) are generally characterized by having larvae with a neph-
ros which terminates 4-14 myomeres anterior to the anus on
the next to last gut loop or between the last and next to last gut
loop (Fig. 57-lower). This condition has not been observed in
any genera of the Ophichthinae outside of the Ophichthus lin-
eage of the Ophichthini. The larvae of Myrichthys, one of the
specialized genera in the Ophichthus lineage, has a nephros which
terminates above or just anterior to the anus (Leiby, in press
a). Blache (1977) identified a series of larvae as Brachysomophis
atlanlicus. This series of larvae differs from the larvae of the
closely related genus Aplalophis in having the nephros termi-
nating above or just anterior to the anus. Larvae of the western
Pacific species of Brachysomophis have not yet been identified.
Consequently, it is unknown whether this nephric position is a
secondarily derived character of the genus Brachysomophis or
whether it is limited to the eastern Atlantic species B. atlanlicus.
The other lineage to arise from the generalized Ophichthus-
hke ancestor contains eight genera including Quassiremus and
Malvoliophis (Fig. 55), which are characterized by various re-
ductions and modifications of the generalized Ophichthus-Vike
condition such as reduced gill arches, cephalic lateralis systems,
and pectoral fins. This lineage probably gave rise to the Sphag-
ebranchini and subsequent lineages by continued modification,
reduction, and specialization of the ophichthin condition
(McCosker, 1977). The larvae of the Quassiremus- Malvoliophis
lineage are virtually unknown. Leiby (in press) tentatively
identified three larvae as Quassiremus produclus, but no other
larvae from this lineage have been identified. There is a natural
LEIBY: OPHICHTHIDAE
107
progression in larval morphology from some Ophichthus spp.
through Quassiremus morphology to sphagebranchin mor-
phology which tends to support McCosker's (1977) hypothesis
that the other ophichthin lineages arose through modification,
reduction, and specialization of the ancestral Ophichthus-like
condition. Quassiremus larvae look much like the larvae of
some Ophichthus spp., but differ in having the nephros termi-
nate over or just anterior to the anus, and in having reduced
gill arches.
The tribe Sphagebranchini is distinguished from the other
tribes of the Ophichthinae by a combination of the following
adult characters: the pectoral girdle is reduced; the pectoral fin
is absent; the gill openings are low to entirely ventral; the neu-
rocranium is elongate (neurocranium depth going 4 or more
times into its length), generally depressed, and truncate poste-
riorly; the gill arches are generally much reduced; the body is
equal to or shorter than the tail; the tail tip is sharply pointed;
and, the cephalic lateralis system is generally better developed
than in other tribes (McCosker, 1977). Larval characters which
distinguish this tribe from other tribes in the Ophichthinae or
which distinguish lineages within the tribe, are reflections of the
adult characters (e.g., reduced gill arches, short gut, dorsal fin
origin) (Leiby, 1982). As yet, there are no independent larval
characters which confirm the monophyletic origin of this tribe
or which confirm the proposed lineages within the tribe, al-
though the larval morphology is similar to, and sometimes dif-
ficult to distinguish from, the larval morphology of some Oph-
ichthini and is consistent with the hypothesis of modification,
reduction, and specialization of the ancestral ophichthin con-
dition which has been proposed based on adult data.
The tribe Bascanichthyini, apparently derived from a mod-
erately specialized ophichthin-like ancestor (McCosker, 1977),
is distinguishable from the other tribes of the Ophichthinae by
a combination of the following adult characters: the body is
equal to, or longer than the tail; the gill openings are low lateral
and crescentic, never entirely ventral; dorsal-fin origin is on the
head in most genera; the pectoral fin is reduced or absent; the
cephalic lateralis system is reduced; and, the gill arches are
generally much reduced (McCosker, 1977). The genus Dalophis
is provisionally placed in the Bascanichthyini despite its pos-
session of a gill arch skeleton and a body length which are more
ophichthin than bascanichthyin, due to its reductions, general
cephalic appearance and several osteological characters (Mc-
Cosker, 1977). If this placement oi Dalophis is correct, it seems
likely that the ancestral bascanichthyin was similar in appear-
ance to Dalophis. Larval characters which distinguish this tribe
from other tribes in the Ophichthinae are reflections of adult
characters (e.g., reduced gill arches, relatively long gut and opis-
thonephros, and dorsal-fin origin). Larvae have been identified
from each of the three proposed bascanichthyin lineages [e.g.,
Dalophis (Blache, 1 977; Palomera and Fortuno, 1981), Bascan-
ichth\'s(B\?Lc\\e, \971\ Leiby, 1981), Gordiichthys (Leiby. in press),
Caralophia (Leiby, in press)], but there are currently no clear
larval characters which are useful for elucidating relationships
within the Bascanichthyini. With one exception, all of the larvae
assigned to the Bascanichthyini are characterized by extremely
low to moderately developed gut loops and, except for gut length,
nephros length and dorsal-fin origin, look much like larvae of
the Sphagebranchini. One larval form which cannot yet be as-
signed to a genus, has tentatively been placed in the Bascani-
chthyini based on gill arch and caudal osteology although its
gut morphology is more like some Callechelyini than Bascani-
chthyini (Leiby, in press). Discovery of the adults of this species
may help clarify relationships within the Bascanichthyini.
The tribe Callechelyini is apparently derived from a bascan-
ichthyin-like ancestor. Adults of this tribe are distinguished by
a short neurocranium (neurocranium depth > 33% of its length);
the dorsal-fin origin on the head or nape; the body longer than
the tail; absence of a pectoral fin; low lateral to entirely ventral
anteriorly convergent gill openings; reduced gill arches; reduced
cephalic lateralis system; laterally compressed body; small eyes;
and, a stout hyoid (McCosker, 1977). Larvae of three of the five
known Callechelyin genera have been identified (Leiby, 1984)
and are readily distinguishable from larvae of the other ophich-
thin tribes. Callechelyin larvae are characterized by moderate
to pronounced gut loops; variable but distinctive pigmentation
(see Leiby, in press b, for full descriptions); anterior dorsal-fin
origin; nephric myomeres more than 56% of total myomeres; a
distinct fourth hypobranchial which may be separate from or
united with a reduced fifth ceratobranchial (a remnant of the
fourth hypobranchial united with a reduced fifth ceratobranchial
may occasionally be found in gill arches of larval Sphagebran-
chini and Bascanichthyini; a distinct fourth hypobranchial is
found in some larval Ophichthini, but, when present, is united
with a well developed fifth ceratobranchial); and usually two
hypurals rather than the three seen in other ophichthids.
McCosker and Rosenblatt (1972) and McCosker (1977) recog-
nized the presence of subgeneric lines in the genus Callechelys.
Evidence from larval morphology confirms the presence of two
subgeneric lineages in Callechelys (Leiby, 1984). Adults of one
subgenus have a split urohyal and two rod-shaped elements in
the pectoral girdle. The larvae of this subgenus have pronounced
gut loops; the fourth hypobranchial free from the fifth cerato-
branchial; most or all of the myosepta without pigment; most
or all of the anal pterygiophores without pigment; no pigment
on the esophagus; pigment on the dorsal surface of each gut
loop but no pigment between gut loops; pronounced, round
pigment patches in the body wall lateral to each gut loop; and,
three to five pronounced, circular postanal pigment patches which
consist of subcutaneous and body-wall pigment. Adults of the
second subgenus have a simple urohyal and one or two rod-
shaped elements in the pectoral girdle. The larvae of this sub-
genus have moderate gut loops; the fourth hypobranchial united
with the fifth ceratobranchial; dark pigment every third to elev-
enth myoseptum, or light pigment on every myoseptum; round
or saddle-shaped patches of pigment in the body wall on the
ventral margin of the tail extending onto the anal pterygio-
phores, or pigment on every anal pterygiophore but none in the
ventral body wall; pigment on the esophagus, on the dorsal
surface of each gut loop, and between each gut loop; occasionally
some body-wall pigment lateral to each gut loop; four to seven
irregular, subcutaneous pigment patches on the tail, usually not
flanked by body-wall pigment.
Relationships to other taxa
The family Ophichthidae is generally considered to be a co-
hesive group which is the sole member of the superfamily Oph-
ichthoidea. The unique nature of ophichthid larvae supports
this allocation. Most workers (e.g., Gosline. 1951; Nelson, 1966b;
McCosker, 1 977) consider the Ophichthidae to be a specialized
offshoot of the Congridae, although Dean (1972) decried the
108
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
value of the characters used to associate the Ophichthoidea with
the Congroidea and implied that the Ophichthidae could just
as easily be a specialized offshoot of the Anguilloidea. While
the only known larvae which could be confused with the
Ophichthidae are members of the family Congridae (e.g., Ac-
romycter larvae have pronounced gut loops. Nystactichthys lar-
vae have a gut which expands abruptly between the esophagus
and intestine), there are no known larval characters which un-
equivocally establish the evolutionary relationships of the
Ophichthidae. Careful osteological studies of ontogenetic series
of eel larvae from the various families may eventually clear the
currently clouded picture.
Department of Natural Resources, Marine Resources
Laboratory, 100 Eighth Avenue Southeast, Saint Pe-
tersburg, Florida 33701.
Clupeiformes: Development and Relationships
M. F. McGowAN AND F. H. Berry
THE order Clupeiformes contains four families of fishes: the
herrings, Clupeidae; the anchovies. Engraulidae; the wolf-
herrings, Chirocentridae; and the denticle herring, Denticipiti-
dae (Nelson, 1976). Denticeps clupeoides. the monotypic den-
ticipitid, occurs in freshwater in southwest Nigeria (Clausen,
1959). Two species of Chirocentrus occur in marine waters of
the Indo-Pacific region from the Red Sea to the western Pacific
(Whitehead, 1972). They are unusual among the Clupeiformes
in that they are piscivorous. The herrings and anchovies are, in
general, small schooling planktivores of marine coastal waters.
The Indo-Pacific shad, Tenualosa reevesii. reaches 509 mm
standard length; the West African riverine species, Thrattidion
noctivagns and Sierrathrissa leonensis. are mature at 18 mm
(Wongratana, 1980). There are 192 species of clupeids in 62
genera and 122 species of engraulids in 16 genera (Table 24)
based on our review of the literature. Herrings and anchovies
are most speciose in the tropics, and individual species are most
abundant in cold temperate regions and eastern boundary cur-
rents (Longhurst, 1971). Some are found in fresh or brackish
water; some are anadromous. They support major fisheries
worldwide. Their biology has been reviewed most recently by
Blaxter and Hunter (1982).
Development
The eggs and the larvae of Chirocentrus are known (Delsman,
1923, 1930b); the egg and larva oi Denticeps are unknown; and
the eggs or larvae of at least one species in a genus have been
described for approximately one-half the genera of herrings and
anchovies but for only one-third of all species. Ontogenetic
stages of herrings and anchovies are best known for species of
commercial interest or potential commercial interest in regions
with low clupeoid diversity such as the northeast Atlantic (e.g.,
Chtpea, Sprattus. Sardina. Engraulis) and the California current
(e.g., Sardinops, Etrumeus. Engraulis). The ontogeny of mor-
phology and behavior, and the requirements for growth and
survival of the herring, Cliipea harengus, and the anchovy, En-
graulis mordax, are well known (Blaxter and Hunter, 1982).
Very little detailed information exists for clupeids from species-
rich areas, especially western African freshwaters and the New
World tropics. Descriptive taxonomy is still needed in these
areas. Table 25 lists the clupeiform fishes for which we found
some information about eggs and larvae.
Published descriptions of clupeoid eggs and larvae may not
be adequate for systematic studies for a variety of reasons. When
there are few species in an area with which to confuse the de-
scribed species, only the key identifying features are described.
When eggs are hatched but the larvae are not reared to meta-
morphosis, usually an atypical starving early larva is described.
When a well-described series of field-caught larvae is compared
with a laboratory-reared series there may be differences in pig-
mentation and size at a particular stage of development due to
the rearing environment. Future descnptions should describe
the eggs and yolk-sac larvae thoroughly because these stages
have characters other than those such as meristics which, be-
cause they are shared with the adults, are redundant for system-
atic purposes. Future descriptions should also try to describe
the development of characters which are of phylogenetic im-
portance in adult-based classifications because the ontogenetic
transformation of a character provides information about the
polarity of states of that character (Nelson, 1978).
Because the eggs and larvae of so many clupeiform genera are
undescribed and because existing descriptions vary in com-
pleteness, it is premature to attempt a phylogenetic classification
of the Clupeiformes based on early life history stages. However,
because many species' eggs and larvae have been described it
is possible to identify and describe characters of taxonomic and
phylogenetic value, to discuss their distribution among the Clu-
peiformes, and to point out some similarities and conflicts be-
tween the distribution of egg and larval characters and current
hypotheses of clupeiform phylogeny.
Taxonomic characters of eggs and larvae
The taxonomic characters of clupeoid eggs include size, shape,
chorion thickness and sculpturing, width of perivitelline space,
degree of yolk segmentation, number and size of oil globules if
present, whether they are pelagic or demersal, whether they are
adhesive or not, and whether they are spawned in fresh, brackish
or full seawater.
The egg of Chirocentrus is 1.60-1.65 mm in diameter, has a
very small perivitelline space, is pelagic, spherical, and is abun-
dant near shore, especially around river mouths (Delsman,
1930b). The egg of Chirocentus nudus has a chorion with fine
hexagonal sculpturing (unique among clupeiforms) and up to 9
small oil globules, while the egg of C. dorab has a smooth cho-
rion and may have a single oil globule (Delsman, 1923, 1930b).
The eggs of clupeids are all globular and they range in size
McGOWAN AND BERRY: CLUPEIFORMES
109
Table 24. Families, Subfamilies, Genera, and Species of Clupeiformes with Selected Meristics. Classification follows Whitehead (1972)
and Nelson (1976) for subfamilies; Wongratana (1980, 1983) and Nelson (1983) where pertinent for genera and species; otherwise the nomenclature
is that of the author cited in the table. Data compiled by F. H. Berry for species presumed valid. A; Atlantic; P: Pacific; c: central; e: east; n:
north; s: south; w: west; FW: Freshwater; IcP: Indo-central Pacific; IwP: Indo-west Pacific; 1: India; Aust; Australia; Philipp: Philippines; US:
United States of America; Braz: Brazil, Venz: Venezuela; Arg; Argentina.
Localion
Dorsal
Anal
P2
Gillrakers
Vertebrae
Upper
Lower
Reference
DENTICIPITIDAE
Denticeps
clupeoides
Nigeria
9
26-27
5
10
41
Clausen, 1959
CHIROCENTRIDAE
Chirocenlrus
dorab
IcP-Aust
72-
-74
Delsman, 1923: White-
head, 1973
nudus
IwP
CLUPEIDAE
Clupeinae
Sardinelta
longiceps
I
17-19
14-18
9
117-241
150-253
Wongratana, 1980
neglecta
se Africa
17-19
16-18
9
108-166
143-188
Wongratana, 1983
lemuru
China-Aust
17-19
15-19
9
51-153
77-188
Wongratana, 1980
Jussieui
China-Aust
19-20
19-21
8
52-61
88-101
Wongratana, 1980
sindensis
I
17-20
17-21
8
16-46
38-77
Wongratana, 1980
gibbosa
IwP
17-20
17-22
8
16-36
38-66
Wongratana, 1980
fimbriata
IwP
18-20
19-22
8
27-47
54-82
Wongratana, 1980
albella
IwP
18-20
18-23
8
20-36
41-68
Wongratana, 1980
dayi
1
18-19
19-20
8
51-103
87-134
Wongratana, 1980
fijiense
N. Guinea
17-18
18-19
8
33-40
61-74
Wongratana, 1980
la Wilis
Philipp
18-19
1-22
Wongratana, 1980
hauliensis
Taiwan
18-20
19-22
8
Wongratana, 1980
brachysoma
1-Aust
17-20
18-22
8
25-39
48-67
Wongratana, 1980
richardsoni
China
18-19
18-22
8
36-42
63-74
Wongratana, 1983
zunasi
China-Japan
17-19
17-21
8
21-23
42-58
Wongratana, 1980
marquesensis
Marquesas
16-18
17-21
7-8
15-58
27-85
42-
-44
Wongratana, 1980
melanura
IcP
16-18
17-20
8
20-41
38-74
Wongratana, 1980
alncauda
se Asia
18-19
17-18
8
20-26
39-43
Wongratana, 1980
aurita
wAeA
17-20
16-18
9
56-81
95-132
45-
-47
Wongratana, 1980
hrasiliensis
wA
17-18
18-20
9
>150
46
Hildebrand, 1963d;
Whitehead, 1973;
Berry
inaderensis
eA
8
>70
Whitehead, 1981
rouxi
ecA
8
34-40
Whitehead, 1981
Amblygasler
sirm
IwP
18-20
17-22
14-18
36-43
Wongratana, 1980
clupeoides
wP
18-19
17-19
12-14
26-31
Wongratana, 1980
leiogaster
IwP
,19
17-20
13-16
31-33
Wongratana, 1980
Herk/olsichlhys
quadrimaculalus
IwP-Aust
18-20
16-21
13-17
30-37
Wongratana, 1980
konigsbergeh
wP-Aust
18-19
19-22
15-17
30-34
Wongratana, 1980
caslelnaui
wP-AusI
17-20
17-22
18-22
39-52
Wongratana, 1980
gotoi
N. Guinea
19
17
16
34
Wongratana, 1983
lossei
Persian G.
18-19
15-18
12-15
29-35
Wongratana, 1983
spilura
I
17-19
15-18
12-15
29-34
Wongratana, 1980
punclatus
Red Sea
17-20
13-18
12-17
31-39
Wongratana, 1980
dispilonotus
se Asia
17-20
16-19
14-17
34-38
Wongratana, 1980
Escualosa
elongala
Thailand
16
19
26
41
Wongratana, 1983
thoracata
IwP-Aust
15-17
17-21
16-25
29-40
Wongratana, 1980
Opisthonema
bidleri
eP
18-21
20-23
8-9
35-47
65-83
46-
-48
Berry and Barrett, 1963
medirasire
eP
17-20
19-23
8-9
70-99
110-156
45-
-48
Berry and Barrett, 1963
herlangai
Galapagos
19-20
19-22
8-9
75-117
133-171
46-
-48
Berry and Barrett, 1963
liherlale
eP
17-20
19-22
8-9
1-149
161-224
44-48
Berry and Barrett, 1963
oglinum
wA
18-22
22-25
8-9
43-60
72-107
45-
-49
Berry and Barrett, 1963
captivai
Colombia A
19-20
18-21
8
(c25-28)
49
Rivas, 1972; Berry
110
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 24. Continued.
Location
Dorsal
Anal
P2
Gillraker
Vertebrae
Upper
Lower
Reference
Harengula
humeralis
wA
18
16
8
13-15
26-29
40-41
Whitehead, 1973; Berry
clupeota
wA
18
18
8
14-16
27-31
41-42
Whitehead, 1973; Berry
jaguana
wA
17-18
17-18
7-8
16-20
31-35
41-43
Whitehead, 1973; Berry
peruana
esP
18-19
15-17
8
15-19
31-51
40-42
Berry
thrissina
enP
16-20
14-17
8-9
9-18
24-33
40-43
Hildebrand, 1946; Berry;
Miller and Lea, 1972
Ramnogaster
arcuata
Arg
7
Whitehead, 1973, 1965
melanostoma
Arg
Whitehead, 1965
pallida
Arg
Whitehead, 1965
Platanichthys
platana
Braz
14
16
7
13
25
Whitehead, 1973
Sardinops
sagax sagax
esP
17-20
17-20
8
49-54
Ahlstrom
sagax caerulea
enP
17-20
17-20
8
21-23
44-45
48-54
Berry; Miller and Lea,
1972
neopilchardus
Aust
18-20
17-21
58-93
50-52
Berry
melanosticta
e Asia
ocellata
s Africa
8
Whitehead, 1981
Sardina
pilchardus
enA
17-18
17-18
8
44-106
50-53
Whitehead, 1981
Rhinosardinia
amazomca
Guyanas
13-16
15-19
8
ca. 20
33-43
Hildebrand, 1963d;
Whitehead, 1973;
Berry
bahiensis
Braz
17
18
Hildebrand, 1963d
Lile
piqmtinga
wcA
15-17
17-19
7-8
12-17
30-36
38-41
Whitehead, 1973; Berry
stolifera
eP
17-18
17-23
8
13-18
32-36
42-44
Hildebrand, 1946
Clupea
harengus
nA
16-20
16-20
37-52
53-60
Hildebrand, 1963d;
Wheeler, 1969
pallasi
nP
13-21
14-20
20
45
46-55
Berry, 1964b, Ahlstrom;
Miller and Lea, 1972
bentincki
Chile
Whitehead, 1965
Sprallus
spraltus
enA
16-19
18-20
7-8
46-49
Whitehead, 1965;
Wheeler. 1969
antipodum
Aust
8
Whitehead, 1965
muelten
Aust
8
Whitehead, 1965
hassensis
Aust
8
46
Whitehead, 1965
fuegensis
Chile
8
49-51
Whitehead. 1965
Clupeonella
cultiventris
Whitehead, 1965
grimmi
Whitehead, 1965
engraulifonnis
Whitehead. 1965
abrau
Whitehead. 1965
Dussumieriinae
Eirumeus
teres
Cosmop.
18-22
10-19
8-9
12-15
28-35
48-50
Wongratana. 1980;
Miller etal.. 1979;
Miller and Lea, 1972
whiteheadi
S. Africa
18-20
12-13
8
16-18
36-39
54-56
Wongratana, 1983
Dussumieria
elopsoides
IcP
18-23
14-18
8
11-16
21-32
54-55
Wongratana, 1980;
Delsman, 1925
acuta
1-China
19-22
14-18
8
11-15
19-26
54-55
Wongratana, 1980;
Delsman, 1925
McGOWAN AND BERRY: CLUPEIFORMES
Table 24. Continued.
Ill
Location
Dorsal
Anal
P2
Gillrakcrs
Venebrae
Upper
Lower
Reference
Spratelloidinae
Spralelloides
gracilis
IwP Aust
11-14
11-14
S
10-12
28-37
Wongratana, 1980
lewisi
N. Guinea
11-13
10-13
8
9-11
28-32
Wongratana, 1983
delkatulus
IwP Aust
10-14
9-11
8
9-12
26-32
44-
-45
Wongratana, 1980;
Miller etal., 1979
robustus
Aust
12-13
10-11
8
9-11
28-35
Wongratana, 1980
Jenkinsia
lamprolaenia
wcA
12-13
13-16
8
19-24
39-
-40
Whitehead. 1973; Berry;
Cervigon and
Velazquez, 1978
stolifera
wcA
9-12
13-16
18-25
Whitehead, 1973
majua
wcA
11-13
21-28
Whitehead, 1973
parvula
Venz
10-13
12-16
20-24
38-
-39
Cervigon and Velaz-
quez, 1978
Dorosomatinae
Clupanodon
ihrissa
wP
16
21-26
8
(190-480)
(200-420)
Wongratana, 1980
Konosirus
punctatus
China
16-19
21-25
8
(145-270)
(160-250)
Wongratana, 1980
Nematalosa
erebi
Aust
14-16
19-22
8
(155-370)
(145-370)
Wongratana, 1980
chanpole
IwP
15-17
22-26
8
(250-315)
(255-355)
Wongratana, 1980
arabica
I
17-19
18-20
8
(145-335)
(180-390)
Wongratana, 1980
come
I-Aust
17-18
20-24
8
(175-245)
(170-250)
Wongratana, 1980
nasus
I-wP
15-19
20-26
8
(155-310)
(165-315)
Wongratana, 1980
japonica
wP
16-18
19-22
8
149-205
156-193
Wongratana, 1980
vlaminghi
Aust
16-17
19-25
8
216-300
239-328
Wongratana, 1980
paubuensis
N. Guinea
14-16
22-27
8
72-342
82-309
Wongratana, 1980
flyensis
N. Guinea
14-16
21-26
8
152-553
195-508
Wongratana, 1983
Gonialosa
whitcheadi
Burma
15
27
8
(92)
90-93
Wongratana, 1983
mammmna
I
14-16
22-27
8
87-160
96-166
Wongratana, 1980
modesta
Burma
15-17
24-28
8
(125-170)
(140-185)
Wongratana, 1980
Anodontostoma
chacunda
IwP
17-21
17-22
8
52-98
54-96
Wongratana, 1980
selangkat
wP
18-20
17-21
8
129-186
100-166
Wongratana, 1980
ihailandiae
IwP
17-20
18-23
8
43-125
46-140
Wongratana, 1983
Dorosoma
cepedianum
wnP
10-13
25-36
7-8
(ca. 300-
400)
48-
-51
Miller, 1960; Berry
petenense
wnA
11-14
17-27
7-8
(ca. 300-
400)
40-
-45
Miller 1960; Berry
anale
eMexico
29-38
Miller, 1960
chavesi
eNicaragua
12-14
(22-31)
Miller, 1960
smithi
wMexico
9-13
(22-31)
43-
-46
Hildebrand, 1963d;
Miller, 1960
Berry
Congothnssinae
Congothrissa
gossei
Congo
14-16
15-17
7-8
ca.
40
Poll, 1964
Alosinae
Hilsa
kelee
IwP
16-19
17-22
8
(45-105)
(70-180)
Wongratana, 1980
Tenualosa
toli
IwP
17-18
15-21
8
(38-55)
(60-95)
Wongratana, 1980
macrura
Java
19
21-22
8
(46-52)
(63-74)
Wongratana, 1980
reevesii
wP
17-19
16-20
8
53-131
80-248
Wongratana, 1980
ilisha
wP
17-20
18-23
8
46-196
62-272
Wongratana, 1980
thibaudeaui
Thailand
16-18
19-23
8
(170-248)
(205-320)
Wongratana, 1980
112
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 24, Continued.
Upper
Gadusia
chapra
variegata
Alosa
sapidissima
pseudoharengus
mediocris
chn^sochloris
alabamae
aestivalis
fallax
alosa
Pakistan
Burma
wnA-enP
eUS-Canada
eUS
eUS
eUS
eUS, Canada
enA
enA
14-18
16-17
17-20
15-19
15-20
16-21
16-20
15-20
18-21
18-21
21-25
25-27
20-23
17-21
19-23
18-21
19-22
16-21
19-23
20-26
(160-235) (170-255)
(250-270) (252-270)
Wongratana. 1980
Wongratana, 1980
59-73
54-59
Hildebrand, 1963d;
Berry
38-43
46-50
Hildebrand, 1963d;
Berry
18-23
54-55
Hildebrand, 1963d;
Berry
20-24
53-55
Hildebrand, 1963d;
Berry
42-48
55
Hildebrand. 1963d;
Berry
41-51
49-53
Hildebrand, 1963d;
Berry
20-40
55-59
Whitehead, 1981;
Wheeler, 1969
55-85
57-58
Whitehead, 1981;
Wheeler, 1969
Ethmalosa
fimbriata
eA
18
22
8
53
136
44
Whitehead, 1981; Berry
Brevoortia
aurea
Braz
gunteri
Gulf Mexico
17-20
20-25
7
144
42-
-44
Hildebrand, 1963d
patronus
Gulf Mexico
17-21
20-23
7
138-142
42-
-48
Hildebrand, 1963d; Berry
smilhi
eUS
18-20
22-23
7
151
45-
-47
Hildebrand, 1963d; Berry
tyrannus
eUS, Canada
18-22
18-24
7
137-145
45-
-50
Hildebrand, 1963d; Berry
Ethmidium
chilcae
Chile-Peru
18-23
15-18
7-8
123-129
147-159
48-
-50
Hildebrand, 1946; Berry
Pellonulinae
Ehirava
fluvial His
I
14-16
12-18
8
12-14
24-30
Wongratana, 1980
madagascarensis
Nelson, 1970
Dayella
malabanca
I
14
17
8
10-11
24-27
Wongratana, 1980
Clupeoides
borneensis
Borneo
15-18
15-19
8
9-12
18-24
Wongratana, 1980
hypselosoma
Borneo
14-15
16-18
8
10
12-19
Wongratana, 1980
paupensis
Borneo
13-16
17-22
8
9-11
15-19
Wongratana, 1980
venulosus
N. Guinea
13-15
20-22
8
Corica
laciniata
Borneo
15-17
13-16 + 2
8
10-13
23-27
Wongratana, 1980
soborna
I
15-16
14-15 + 2
8
9-11
19-21
Wongratana, 1980
Pellonulinae
Laevisculella
dekimpet
Nelson, 1970
Odaxothrissa
losera
Nelson, 1970
Potamothrissa
aculiroslris
Nelson, 1970
Spratellomorpha
bianalis
Nelson, 1970
Pristigasterinae
llisha
sirishai
I
17-18
39-43
8-12
22-26
Wongratana, 1980
novacula
Burma
16
43-45
10-12
21-23
Wongratana, 1980
megaloplera
1
16-19
38-53
8-11
19-23
47-
-48
Wongratana, 1980; Berry
elongala
1-China
16-20
43-53
9-13
21-25
Wongratana. 1980
filigera
I
17-21
46-52
9-12
19-23
50-
-52
Wongratana, 1 980; Berry
macrogaster
I
18-19
49
11-12
23-25
Wongratana, 1980
pristigaslroides
Java
17-18
47-48
9-10
17
Wongratana, 1980
kampeni
1
16-18
38-46
9-12
20-24
Wongratana, 1983
striatula
1
15-18
40-48
10-13
21-24
Wongratana, 1980
melastoma
IwP
15-18
35-48
10-13
21-25
Wongratana, 1983
McGOWAN AND BERRY: CLUPEIFORMES
113
Table 24. Continued.
Location
Dorsal
Anal
P2
Gillrakcrs
Venebrae
Upper
Lower
Reference
obfuscala
I
16
39-42
7
12-13
27-28
Wongratana, 1980
afncana
ecA
15
47
Whitehead, 1981
amazonica
Braz
20
34
6
15
29
Hildebrand, 1963d
furlhii
ecP
15-17
46-50
11-12
20-25
50-52
Peterson, 1956;
Hildebrand, 1946;
Meek and Hildebrand,
Neoopisthoplerus
1923
cubanus
Cuba
12-15
39-43
0
10
17-19
47
Hildebrand, 1963d, Berry
tropicus
15
43-48
0
8
20
45-47
Peterson, 1956;
Hildebrand, 1946
Pellonulinae
Clupeichthys
hleekeh
Borneo
14-15
16-18 + 2
8
8-10
16-18
Wongratana, 1980
aesarnensis
Thailand
13-15
14-16 + 2
8
8-10
17-19
Wongratana, 1983
goniognathus
Thailand
14-15
15-17 + 2
8
8
15-16
Wongratana, 1980
perakensis
Malaya
13-15
14-17 + 2
7
5-9
13-15
Wongratana, 1980
Pellonula
leonensis
ecA
8
20-30
Whitehead, 1981
vorax
ecA
Whitehead, 1981
Microthrissa
royauxi
Nelson, 1970
Poecilothrissa
congica
Nelson, 1970
Hyperlophus
villala
Nelson, 1970
Cynolhnssa
ansorgii
Whitehead, 1981
memo
Potamalosa
richmondia
Wongratana, 1980
Gitchnstella
aestuarius
Wongratana, 1980
Limtwlhrissa
mtodon
Wongratana, 1980
Stolothrissa
tanganicae
Wongratana, 1980
Pristigasterinae
Prist igaster
cayana
Brazil
13-16
44-55
0
10
20-23
43-44
Hildebrand, 1963d; Berry
Opisthoplerus
valenaermesi
China
16-18
54-65
7
9-12
23-25
Wongratana, 1980
lardoore
I
14-17
51-63
7
8-12
22-28
50-52
Wongratana, 1980; Berry
dovii
ecP
12-13
53-62
0
17-18
51-52
Meek and Hildebrand,
1923; Ahlstrom
equalorialis
esP
11-12
59-62
0
10
25
46-47
Hildebrand, 1946;
Ahlstrom
Raconda
russehana
I
0
81-92
8-11
23-27
62
Wongratana, 1980; Berry
Pellona
ditchela
I-Aust
16-19
34-41
7
10-14
22-27
42
Wongratana, 1980; Berry
day!
I
17-18
35-42
7
9-11
20-21
Wongratana, 1983
altamazonica
Braz
18
37-40
6-7
9
12-14
Hildebrand, 1963d; Berry
castelnacana
Braz-Venz
18-20
34-42
6-7
13-14
24-25
45-46
Hildebrand. 1963d;
Whitehead, 1973; Berry
flavipinnis
Braz-Arg
17-21
38-47
7
14-15
28-31
43
Hildebrand, 1963d;
Whitehead, 1973; Berry
harroweri
wcA
14-17
36-42
5-6
12-13
24-28
38-40
Hildebrand, 1963d;
Whitehead, 1973; Berry
114
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 24. Continued.
Location
Dorsal
Anal
P2
Gillrakers
Vertebrae
Upper
Lxjwer
Reference
Odontognathus
mucronatus
wsA
10-12
74-
-85
0
7-9
22-26
53-
-54
Hildebrand, 1963d;
Whitehead, 1973; Berry
compressus
wcA
10-14
52-
-62
0
9
18-23
46-
-47
Hildebrand, 1963d;
Whitehead, 1973; Berry,
Meek and Hildebrand,
1923
panamensis
ecP
11-12
61-68
0
ca. 21
51-
-53
Peterson, 1956;
Meek and Hildebrand,
Chirocentrodon
1923
bleekenanus
wcA
14-16
38-
-45
6-7
4-6
15-17
44-
-45
Hildebrand. 1963d;
Whitehead, 1973; Berry
Pliosteostoma
lutipinnis
ecP
49-51
0
50-51
Peterson, 1956; Berry
macrops
CLUPEIDAE
Status not verified
Alosinae
Caspialosa
maeolica
Nelson, 1970
Clupeinae
Clupeonella
delicalula
Nelson, 1970
Dorosomatinae
Nematatosa
horm
Nelson, 1970
Thratlidion
noctivagus
Sierrathrissa
leonensis
ENGRAULIDAE
Coilinae
Coilia
ramcarati
I
14-16
9-10
21-23
29-30
Wongratana, 1980
borneensis
Borneo
14-15
7
21-23
32
Wongratana, 1980
reynaldi
I
13-14
7
20-27
28-36
Wongratana, 1980
coomansi
Borneo
14
7
21-24
31-33
Wongratana, 1980
rebentischii
Borneo
14-15
7
15-19
22-27
Wongratana, 1980
neglecta
I
13-15
7
17-19
21-27
Wongratana, 1980
dussumieri
I
13-15
7
17-20
23-26
Wongratana, 1980
rendahli
China
13-15
7
grayii
I-China
13-14
7
21-23
28-31
Wongratana, 1980
lindmam
Thailand
12-15
7
18-25
29-34
Wongratana, 1980
macrognalhos
Borneo
14-15
7
15-16
22-24
Wongratana, 1980
mystus
China
13-15
79-
-89
6-7
17-22
25-29
Wongratana, 1980
nasus
China-Japan
13-15
87-
-117
7
16-20
23-26
Wongratana, 1980
Engraulinae
Engraulis
japonicus
IwP
14-17
14-22
22-34
26-39
Wongratana, 1980
(=australis)
eA
(=encrasicolus)
eA
Wongratana, 1980
(=capensis)
sAfrica
Wongratana, 1980
anchoita
swA
Whitehead, 1973
euryslole
nwA
15-16
16-
-19
7
28-31
43-
-45
Whitehead. 1973
ringens
seP
15-18
19-
-24
35-43
38-48
46-
-49
Hildebrand, 1946; Berry
mordax
neP
14-19
19-
-26
28-41
37-45
43-
-47
Miller and Lea. 1972
"juruensis"
Amazon
Whitehead, 1973
A nchovia
clupeoides
swA
14
31
7
105
41
Whitehead, 1973
rastralis
eP
12-14
26-
-30
ca. 50
Meek and Hildebrand,
1923; Whitehead, 1973
tnnilatis
cubana
parva
lamprotaenia
hepselus
filfera
lyok'pis
ginsburgi
tricolor
choerosloma
januaria
mitchilli
pecloralis
cayorum
argenteus
argentivitlala
ischana
McGOWAN AND BERRY: CLUPEIFORMES
Table 24. Continued.
115
Location
Dorsal
Anal
P2
GUItakei^
Vertebrae
Upper Lower
Reference
surinamensis
macrolepidota
magdalenae
cwA
eP
neP
13-15
12-14
25-28
27-33
7
47-62
ca. 95
40-42
Whitehead, 1973
Meek and Hildebrand,
1923; Whitehead, 1973;
Peterson, 1956
A nchoa
spinifer
wcA-ecP
15-17
30-40
7
12-16 12-18
19-21 ±
Hildebrand, 1963c;
Venz
wA
wcA
wA
wA
wcA
wA
Venz
pananwnsis
ecP
compressa
mundeoloides
walkeri
anatis
curta
ecP
delicatissima
P
helleri
P
slarksi
ecP
clarki
eigenmanma
P
ecP
scofteldi
P
lucida
ecP
13-15
14-15
26-32
14-16
20-24
15-16
21-25
13-16
19-27
13-16
18-24
13-15
19-23
12-16
19-27
18-22
wsA
14-16
18-22
Bermuda
13-15
22-24
wsA
14-15
21-24
wnA
14-16
24-30
Braz
14-16
25-27
wA
13-15
25-29
Venz
16
32
ecP
18-20
enP
18-21
ecP
12
32-26
14-19
19-21
16-18
23-25
13-15 23-26 18-21 23-26
20-23
32-40
20-22
21-22
41
7
7
17-23
17-20
23-33
23-28
42-43
38-41
7
13-18
16-22
39-42
7
15-21
19-25
40-44
7
17-19
20-26
39-40
7
16-23
20-27
41-43
44-45
Whitehead, 1973;
Peterson, 1956;
Cervigon, 1966;
Nelson, 1983
Whitehead. 1973;
Cervigon. 1966;
Hildebrand, 1963c
Whitehead, 1973
Whitehead, 1973;
Hildebrand, 1964
Whitehead. 1973;
Hildebrand, 1964
Whitehead, 1973;
Hildebrand, 1964
Whitehead, 1973;
Hildebrand, 1964
Whitehead, 1973;
Cervigon, 1966;
Hildebrand, 1963c
Cervigon, 1966;
Hildebrand, 1963c
25-28
18-22
24-28
40-
-42
Hildebrand. 1963c
17-20
23-26
41-
-42
Hildebrand, 1963c
20-23
23-26
41-
-42
Hildebrand, 1963c
15-19
20-26
38-
-44
Hildebrand, 1963c
13-14
17-19
4
2
Hildebrand, 1963c
13-15
15-17
43
Hildebrand, 1963c
14
19
Hildebrand, 1963c
17-21
24-25 +
19-22
Peterson, 1956; Nelson,
1983
19-21
22-24 +
19-21
Peterson, 1956; Nelson,
1983
22-24
18-20 +
21-24
18-19 +
20-22
18-20 +
21-23
18-20 +
21-24
17-19 +
20-23
Peterson, 1956; Nelson.
1983; Hildebrand,
1946
Nelson, 1983
Nelson, 1983
Nelson, 1983
Nelson, 1983
22-25
19-22 +
19-22
Peterson, 1956; Nelson.
1983
23-26
19-21 +
19-21
20-23 +
18-21
Nelson. 1983; Miller
and Lea, 1972
Nelson, 1983; Miller
and Lea, 1972
22-26
20-22 +
19-21
21 + 21
Peterson, 1956; Nelson,
1983
Nelson, 1983
12-13
17-21 +
20-25
20-22 +
21-23
Peterson, 1956; Nelson,
1983
Nelson, 1983
19-22
17-20 +
19-22
Peterson, 1956; Nelson,
1983
116
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 24. Continued.
Location
Dorsal
Anal
Gil I rakers
Vertebrae Refere
'2 Upper
Lower
ice
naso
ecP
14-16
23-27 21
-24 23-27
19-21 +
19-22
Peterson, 1956; Nelson,
1983; Hildebrand,
1946
chamensis
eP
21 + 22
Nelson, 1983
nasus
ecP
15-16
21-27 21
-25 24-28
20-21 +
20-22
Nelson, 1983;
Hildebrand,
1946
exigua
ecP
17-22
23-25
43-45
Peterson, 1956; Nelson,
1983
Anchovielta
leptdentostole
wsA
14-16
22-25
7 17-18
19-23
Whitehead, 1973;
Cervigon, 1966
urevirostris
wsA
16-18
18-20
7
24-27
Whitehead, 1973
guianensis
wcsA
14-15
18-20
7
23-26
40 Whitehead, 1973
cayennensis
wcA
13
16
7
30
Whitehead, 1973
nattereri
Braz
12
25-29
Whitehead, 1973;
Cervigon, 1966
perfasciata
wnA
12-15
15-19
18-23
25-28
42-44 Cervigon, 1966
elongata
Panama A
13-14
22-24
17-18
22-24
39 Cervigon, 1966
blackbumi
Venz
13-15
25-27
10-12
15-17
43 Cervigon, 1966
jainesi
Braz
12-13
19-21
12-13
20-21
40 Cervigon, 1966
vaillanti
23
19
Whitehead, 1973
carrikeri
17-18
14-15
Whitehead. 1973
Slolephorus
indicus
IwP
14-17
17-22
16-20
20-28
20-23+ Wongratana,
19-21
980
commersonii
IwP
15-17
20-23
12-27
21-35
Wongratana,
980
brachycephalus
Papua
16-17
22-25
15-17
20-22
Wongratana,
983
chinensis
China
16-18
20-23
18-19
26-27
Wongratana,
980
wailei
1-Aust
15-17
19-24
14-17
1-4
Wongratana,
980
holodon
seAfr
15-18
20-23
17-22
24-29
Wongratana,
980
andhraensis
el-Papua
15-17
19-23
14-15
20-21
Wongratana,
980
lysoni
Papua
15-17
21-25
15-18
21-25
Wongratana.
983
insulahs
I-China
14-17
19-23
16-20
22-28
Wongratana,
980
dubwsus
I
14-16
19-24
19-24
25-31
Wongratana,
980
baganensis
I
14-16
20-23
16-19
20-24
Wongratana,
980
iri
Thailand
14-15
19-22
15-17
19-22
Wongratana,
980
oligobranchus
Philipp
14-16
18
7 13-14
17-18
Wongratana,
983
Thryssa
baelama
IwP
15
29-34
14-20
19-26
Wongratana,
980
chefuensis
China
14
29-34
23-28
27-30
Wongratana,
980
rastrosa
N. Guinea
14-15
32-35
39-44
55-61
Wongratana,
980
scratchteyi
N. G.-Aust
14
33-36
15-18
18-20
Wongratana,
980
aesluaha
N. G.-Aust
13-15
32-36
22-25
27-29
Wongratana,
980
kammalcnsis
Thailand
14-15
32-37
23-27
28-32
Wongratana,
980
kammalensoides
I
14
34-35
18
24-25
Wongratana,
983
vilrirostris
e Africa
13-15
34-43
14-17
20-23
Wongratana,
980
adetae
China
13-14
38-44
13-16
20-22
Wongratana,
980
dussumieri
I-Taiwan
12-15
34-38
13-16
17-19
Wongratana,
980
mysto-x
I-China
13-15
35-39
9-11
13-16
Wongratana,
980
polybranchialis
I
13-15
38-42
18-21
25-27
Wongratana,
983
gualamiensis
I
13-15
36-40
11-13
17-19
Wongratana,
980
malabarka
I
13-15
37-41
14-16
17-19
Wongratana,
980
hamiltonii
IwP
13-15
35-41
7-10
11-15
Wongratana,
980
whiteheadi
Pers. G.
12-14
42-46
13-15
18-20
Wongratana,
983
purava
I
12-14
42-47
14-16
18-19
Wongratana.
980
stenosoma
I
12-14
43-48
13-15
17-19
Wongratana,
983
dayi
I
13-14
44-49
10-13
14-18
Wongratana,
983
spinidens
I-Thai
12-14
44-48
9-11
13-15
Wongratana,
980
setirostris
I-China
13-15
32-39
5-6
10-12
Wongratana,
980
Encrasicholina
purpurea
Hawaii
12-15
14-18
7 1 5-22
23-29
41-44 Miller etal., 1
Wongratana
Nelson, 198
979;
, 1980;
3
McGOWAN AND BERRY: CLUPEIFORMES
Table 24. Continued.
117
Localion
Dorsal
Anal
P2
Gillrakers
Vertebrae
Upper
Lower
Reference
punclifer
IwP
12-16
14-17
7
15-22
23-29
24-25 +
17-20
Miller etal., 1979;
Wongratana, 1980;
Nelson. 1983
heterolobus
IwP
13-15
15-19
20-25
23-29
22-24 +
19-21
Miller etal., 1979;
Wongratana, 1980;
Nelson, 1983
devisi
I-Aust
13-16
17-21
17-18
20-22
21-23 +
19-21
Miller etal.. 1979;
Wongratana, 1980;
Nelson, 1983
ronquilloi
Philipp
15-17
19-22
20-21
28-30
Wongratana, 1980
Pterengraulis
alherinoides
wcA
12-14
29-35
7
10-12
12-15
43-45
Lycengraulis
hatesii
wcA
14-16
27-30
7
9-13
12-15
47
Whitehead, 1973;
Cervigon, 1966
grossidens
wcA
14-16
24-28
7
13-19
17-23
41-48
Whitehead, 1973;
Cervigon, 1966
poeyi
eP
13-15
22-27
14-18
18-23
43
Whitehead, 1973;
Peterson, 1956;
Meek and Hildebrand, 1923
Cetengraulis
edenlulus
wcsA
13-16
21-27
7
45-53
Whitehead, 1973;
Meek and Hildebrand, 1923
mysticelus
ecP
13-17
18-26
40-58
43-60
39-43
Peterson, 1956;
Hildebrandichthys
seliger
Venz
12
25
Papuengraulis
micropinna
N. Guinea
5-6
54-56
Lycolhrissa
crocodilus
China
10-13
47-51
Setipinna
tenuifilis
papuensis
melanochir
taty
wheeleri
phasa
brevifilis
I-China
N. G.-Aust
China
I-China
Burma
I
I
14-16
14-15
13-15
13-15
14
13-15
13-15
49-59
54-57
48-53
48-58
72-77
69-82
68-75
Heterothrissa
breviceps
I-China
17-18
59-64
Status not verified:
Thrissa
grayi
Lycengraulis
barboun
olidus
Cetengraulis
juruensis
Amazon-FW
20-22
Anchoa
arenicola
Anchoviella
hubbsi
pallida
balboae
llisha
indica
ca. 23
ca. 33
6-7
15-16
25-27
6-7
8-10
10-12
13-17
11
15
7-10
9-12
13-17
18-20
16-18
21-22
15-16
18-19
14-15
17
7-8
11-12
Hildebrand, 1946; Meek
and Hildebrand, 1923;
Miller and Lea, 1972
Cervigon, 1966;
Schultz, 1949
Wongratana, 1980
Wongratana, 1980
Wongratana, 1980
Wongratana, 1980
Wongratana, 1980
Wongratana, 1980
Wongratana, 1983
Wongratana, 1980
Wongratana, 1980
Wongratana, 1980
Nelson, 1970
Nelson, 1970
Nelson, 1970
20 + 20 Nelson, 1984
Nelson, 1970
Nelson, 1970
Nelson, 1970
Nelson, 1970
Nelson, 1970
118
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
from 0.59-0.75 mm in Sardinella jussiem (Bensam, 1970) to
2.5-3.8 mm in Alosa sapidissima (Jones et al., 1978). Most
clupeid eggs are 1-2 mm in diameter. All have a segmented
yolk. The chorion is not ornamented or sculptured. The peri-
vitelline space varies in thickness among species. It may be as
large as 45% of the egg diameter (Sardinella zunasi) or as small
as 5-10% (Anodontostoma. Opisthoplerus). The egg yolk may
shrink relative to the egg diameter when preserved (Bensam,
1967) and the yolk decreases in size during the development of
the embryo. Oil globules are present in the eggs of most clupeids.
One is often present (e.g.. Sardinella. Harengula, Sardinops);
Escualosa thoracata has nine (Delsman, 1932a, described as
Clupeoides Hie). The eggs of clupeids which lay demersal adhe-
sive eggs (Clupea. Dorosoma, Spratelloides) have a gelatinous
covering around the egg. The pelagic egg of Tenualosa ilisha is
also covered by a gelatinous sheath. In Dorosoma petenense the
adhesive layer is composed of transformed ovarian follicular
epithelium, an unusual feature among teleosts (Shelton, 1978).
Eggs of anchovies, family Engraididae. range in size from 0.7
mm (Lycengraidis) to 1.75 mm (Slolephorus, long axis). Their
shape varies from globular to extremely elliptical. The ratio of
the long axis of the ellipse to the short axis has been used to
identify anchovy eggs (Peterson, 1956; Phonlor, 1978). Some
Slolephorus species have a distinct knob on one end of the egg
surrounding the micropyle. A perivitelline space is present but
smaller and less noticeable than in clupeid eggs because of the
elliptical shape. Oil globules are absent except in the genera
Coilia and Setipinna, which have spherical eggs like clupeids,
and the Indo-Pacific species of Slolephorus. Fig. 58 illustrates
representative eggs of clupeiforms.
Yolk-sac larvae are characterized by their size at hatching (2-
5 mm), which is related to yolk size; whether the yolk-sac is
rounded or pointed posteriorly, the number and position of oil
globules, number of myomeres and pigmentation. Larvae from
demersal adhesive eggs may hatch with pigmented eyes (Clupea
harengus); those from pelagic eggs hatch with unpigmented eyes.
Oil globules may be present in the anterior, ventral, or posterior
part of the yolk sac. Multiple oil globules in early embryos
coalesce into a single large one before hatching in Seiipinna
phasa (John, 1 95 1 a). A spherical yolk sac usually remains spher-
ical although shrinking in size during development (Sardinella
zunasi), while a yolk sac which is pointed posteriorly may be-
come more rounded as yolk is utilized (Coilia sp.). Larval clu-
peiforms are slender and elongate with long straight guts. Series
of melanophores are variously arranged above and below the
gut and along the ventral body wall. Subtle differences in pig-
mentation are very useful for identifying co-occuring larval clu-
peoids prior to fin development. Larvae of Engraulis mordax.
Sardinops sagax. and Etrumeus leres are illustrated for com-
parison in Moser (1981). Median dorsal melanophores in clu-
peid embryos migrate, reaching their characteristic ventral po-
sitions soon after hatching (Orion, 1 953a). In engraulids, pigment
cells are presumed to migrate similarly but they don't become
pigmented until after hatching. Melanophores are commonly
present ventrally just anterior to the pectoral symphysis in small
larvae, (e.g., Opislhonema. Harengula, Engraulis, Sardinops,
Etrumeus). During development external rows of melanophores
become dark streaks and internal melanophores may increase
in size and number at first but disappear or become occluded
at transformation. A thorough description of pigment devel-
opment of laboratory-reared Opislhonema oglinum larvae com-
plete with dorsal, lateral, and ventral illustrations is given by
Richards et al. (1974). Preanal myomere number is taxonom-
ically useful but it does not correspond exactly with precaudal
vertebral count in the adult because of changes during trans-
formation. Pectoral fin buds and a continuous dorsal-caudal-
anal finfold are present at hatching. Fin rays first appear in the
caudal fin then in the dorsal, then the anal, next the pelvic, and
last the pectoral fin. Ossification of fin rays proceeds in the same
order. A full complement of fin rays is not attained until trans-
formation, which occurs at approximately 20 mm standard length
(e.g., Harengula jaguana. Houde et al., 1974; Opislhonema og-
linum Richards et al., 1974). Figs. 59 and 60 illustrate yolk sac
larvae of herrings and anchovies.
The most useful single character for identifying larval clu-
peiforms is total myomere or vertebral number. Pigment pat-
terns are useful when vertebral counts overlap. The relative
positions of dorsal and anal fins and the length of the gut can
be used to separate clupeids from engraulids: clupeids have a
longer gut relative to body length and there is a gap between
the posterior margin of the dorsal fin and the anterior margin
of the anal fin; engraulids have a shorter gut and tend to have
the posterior margin of the dorsal over the anterior insertion of
the anal fin. The number of myomeres between dorsal and anal
fins has been used as a taxonomic character in larvae of certain
size classes (Houde and Fore, 1973) and in clupeid adults (Sve-
tovidov, 1963). During metamorphosis the position of the gut
and the dorsal and anal fins shift forward relative to myomere
number. The dorsal insertion moves 10 myomeres forward in
Sardinops sagax (Ahlstrom, 1968); it moves eight myomeres
in Harengula jaguana (Houde et al., 1974). The migration of
the fin takes place at approximately the time when the fin ray
number stabilizes. The pelvic fin migrates posleriad in Clupea
harengus (Lebour, 1921). Because of these dramatic changes in
morphology during development different characters must often
be used at different stages to separate species. However some
morphometric characters show a small but consistent difference
between species at all sizes as between .4losa pseudoharengus
and .4. aestivalis (Chambers et al., 1976). Additional care must
be taken when using information from laboratory-reared spec-
imens to identify field samples. Fin development began at 4
mm in laboratory-reared Opislhonema oglinum. but was not
observed in wild-caught larvae less than 7 mm long (Richards
et al., 1 974). Shrinkage due to preservation and handling (Thei-
lacker, 1980a) also presents problems when comparing devel-
opment of larvae based on length. Meristic characters in Clupea
Fig. 58. Eggs of Clupeiformes illustrating taxonomic characters: number and size of oil globules, width of perivitelline space, degree of yolk
segmentation, shape, size. (A) Chirocemrus nudus. 1.56 mm. Delsman, 1923; (B) Etrumeus leres. 1.35 mm, Ahlstrom and Moser. 1980; (C)
Opisthoplerus tardoore, 0.85 mm, Bensam, 1967; (D) Dussumiena. 1.5 mm, Delsman, 1925; (E) .Anodontostoma chacunda. 0.92 mm, Delsman,
1926c; (F) Sardinops melanosticta. 1.60 mm, Mito, 1961; (G) Coilia. 1.04 mm, Delsman. 1932b; (H) Setipinna phasa. 1.10 mm, Jones and
Menon, 1950; (I) Anchoa mitchilli. 0.84 x 0.65, Kuntz, 1914b; (J) Engraulis mordax. 1.40 x 0.74, Bolin, 1936a; (K) Slolephorus msulans. 1.92
X 0.69, Delsman, 1931; (L) Slolephorus indicus or commersonii. 1.15 x 0.81, Delsman, 1931. All redrawn by J. Javech.
McGOWAN AND BERRY: CLUPEIFORMES
,19
H
K
120
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 59. Yolk-sac larvae of Clupeidae and Chirocentrus illustrating taxonomic characters: number, size, and position of oil globules; shape of
yolk sac; degree of segmentation of yolk; preanal myomeres. (A) Sardinella zunasi. 2.1 \ mm, Takita, 1966; (B) Sardmelta :unasi, 4.79 mm,
Takita, \9(>(>.(C) Elrumeus teres. AM mm. Mao, \9(i\:(D) llisha elongata. 5.59 mm, Sha and Ruan, \9i\:{E) Dussumieria. 3.17 mm, Delsman,
1925; (F) Chirocentrus mtdus. 3.79 mm, Delsman, 1923. All redrawn by J. Javech.
McGOWAN AND BERRY: CLUPEIFORMES
121
Fig. 60. Yolk-sac larvae of Engraulidae illustrating taxonomic characters: oil globules, shape of yolk sac, yolk segmentation, preanal myomeres.
(A) EngrauUs japomcus. 3.02 mm, Mito, 1961; (B) Coilia. 2.83 mm, Takita, 1967; (C) Coilia. 2.46 mm, Delsman, 1932b; (D) Slolephorus
msularis. 2.19 mm. Delsman, 1931; (E) Thryssa hamiltomi. 2.42 mm, Delsman, 1929a; (F) Cetengraulis mysticetus, 1.99 mm, Simpson, 1959.
All redrawn by J. Javech.
122
ONTOGENY AND SYSTEM ATICS OF FISHES- AHLSTROM SYMPOSIUM
harengus larvae were shown to be affected by temperature and
salinity (Hempel and Blaxter, 1961); morphometric characters
in Gikhristella aestuarius adults were found to differ between
estuaries with different types of prey items (Blaber et al., 1981).
There are several characters which may be useful in system-
atics when they are described for more clupeiform species. The
melanophores on the caudal fin dorsal and/or ventral to the
notochord tip in small larvae have been described for a few
species. Harengida jaguana has dorsal melanophores only at
first, then both dorsal and ventral (Houde et al., 1974). Opis-
thonema oglinum has ventral ones (Richards et al., 1974). Sar-
dinella brasiliensis. S. maderensis and S. zunasi have just ven-
tral melanophores but Sardine/la rouxi has both. Slight
differences in pigmentation over the brain and on the mid-dorsal
and mid-ventral postanal body midline have been used to iden-
tify scombrid larvae. Small scombrid larvae are otherwise very
similar to each other as are clupeoid larvae. The development
of free neuromasts and the lateral line has been described for a
few species (Blaxter et al., 1983). Development of the swim-
bladder and its unique connection with the inner ear should be
useful (Hoss and Blaxter, 1982). Ephemeral basihyal teeth were
observed on Opisthonema oglinum and Harengula jaguana lar-
vae (Richards et al., 1974; Houde et al., 1974). Two patterns
of nasal epithelium cells have been observed with scanning elec-
tron microscopy (Yamamoto and Ueda, 1 978). Harengula, Sar-
dinops and Konosirus had one pattern while Etrumeus (a clu-
peid) had the same pattern as Engraulis, an engraulid.
Although the eggs and yolk-sac larvae of clupeiforms have
many characters of potential systematic importance, the taxo-
nomic characters of the older larvae (meristics, fin position, and
pigmentation) will tend to be redundant with the same adult
characters. However, clupeoids are easily reared in the labo-
ratory so direct experimental evaluation of the polarity of adult
character states by comparative developmental studies is pos-
sible.
Relationships
The clupeiform fishes are considered a well-defined mono-
phyletic group based on their unique otophysic connection, the
caudal skeleton, and other characters (Greenwood et al., 1966).
The distribution of species within genera, genera within subfam-
ilies, and number and taxonomic rank of categories within the
group are not agreed upon (Gosline, 1971, 1980; Miller, 1969;
Nelson, 1967, 1970, 1973; Whitehead, 1972, 1973). J. S. Nelson
(1976) lists the families Chirocentridae, Denticipitidae, Clu-
peidae, and Engraulidae. He gives seven subfamilies of herrings
(Dussumieriinae, Clupeinae, Pellonulinae, Alosinae, Doroso-
matinae, Pristigasterinae, and Congothrissinae) and two
subfamilies of anchovies (Engraulinae and Coilinae). Spralel-
loides is separated from the Dussumieriinae and given subfamily
rank by Whitehead (1972. 1973). Jenkinsia is the western At-
lantic spratelloidin.
Based on the gill arches Nelson (1967) concluded that the
Dussumieriinae (including Spratelloides and Jenkinsia) were the
most primitive clupeid family; the Pristigasterinae were also
primitive but with distinctive specializations; the Clupeinae were
more advanced, but linked to the Dussumieriinae by Clupea
and Sprattus; the Alosinae and Dorosomatinae were closely
related and perhaps both derived from the Clupeinae; and the
Pellonulinae, lacking the specializations of the Alosinae and
Dorosomatinae, most resembled the Clupeinae. Expanding his
study of gill arches in the Clupeidae to the hyobranchial ap-
paratus in the Clupeiformes, Nelson (1970) divided the order
into the superfamilies Chirocentroidae, Engrauloidae. Pristi-
gasteroidae, and Clupeoidae. The Clupeoidae were suggested to
consist of two families: the Clupeidae composed of the Dus-
sumieriinae, Pellonulinae, and Alosinae in part; and the Do-
rosomatinae composed of the Dorosomatinae plus Hilsa from
the Alosinae and Harengida and Sardinella from the Clupeinae.
Sardina and Alosa were aligned with Clupea, Polamalosa, and
Etrumeus in his tree depicting relationships of representative
genera (Nelson, 1970: Fig. 1 1).
Whitehead (1972, 1973) acknowledged that radical changes
in clupeid classification could be expected but retained the
subfamilies Dussumieriinae, Spratelloidinae, Clupeinae, Pel-
lonulinae, Alosinae, Dorosomatinae, and Pristigasterinae in his
works which were chiefly concerned with the identification of
genera and species.
The most recent comprehensive work is that of Wongratana
(1980) on the Clupeidae and Engraulidae of the Indo-Pacific.
He examined over 14,000 specimens and considered many me-
ristic and morphological characters including gill rakers, epi-
branchial organs, predorsal bones, caudal osteology, circumor-
bital bones, gut form, the gas bladder, scale striae, and the patterns
of scale distribution on the body. No numerical, cladistic, or
phenetic analyses were done. Taxonomic characters were dis-
cussed with respect to apparent evolutionary trends and relative
importance. Wongratana retained the subfamilies of Whitehead
(1972). The Spratelloidinae were diagnosed by a bony process
on the 6th and 1 2th principal caudal rays. Spratelloides is also
unique among Indo-Pacific clupeids in having a single epural.
Jenkinsia, the spratelloidin in the Western Atlantic, also has a
single epural (Hollister, 1936). The Alosinae and Dorosomatin-
ae were kept separate and the Pristigasterinae were accorded
subfamily status although considered quite distinct from the
other clupeids. The Dussumieriinae and Pellonulinae were con-
sidered the most primitive groups, the Alosinae and Doroso-
matinae the most advanced, and the Spratelloidinae and Clu-
peinae were considered intermediate. Within the anchovies, the
Coiliinae have one epural while the Engraulinae have two {En-
graulis) or three (Papuengraulis). The Coiliinae were considered
primitive relative to the Engraulinae although specialized in
many respects.
Wongratana ( 1 980) found that the number of predorsal bones
varies from one to thirty in the clupeids and engraulids (Chi-
rocenlrus has none). Some engraulids and pellonulins have a
gap between the posterior predorsal bone and the first dorsal
pterygiophore which he interpreted as evidence that the dorsal
fin has migrated posteriad during evolution. It would be inter-
esting to compare the patterns of dorsal bones and the anteriad
migration of the dorsal fin during larval metamorphosis. The
"dorsal scutes" of Clupanodon ihrlssa were found to be the
exposed tips of predorsal bones (Wongratana, 1980). The only
double-armored herrings known now are Polamalosa and Hy-
perlophus in the Pellonulinae, and Elhmidium in the Alosinae.
Dorsal scutes are interesting because they occurred in herring-
like fossils (Diplomystus, Knightia, and Gasteroclupea) which
resemble pristigasterins (Nelson 1970).
Because he examined so many species from such a wide area
Wongratana (1980) was able to clear up many nomenclatural
questions and to correct previous misidentifications which had
been based on limited material. He also described 24 new species
McGOWAN AND BERRY: CLUPEIFORMES
123
(Wongratana, 1 983) and provided keys to all Indo-Pacific species
(Wongratana, 1980). However no direct comparison between
his classification and that of Nelson (1967, 1970, 1973) is pos-
sible because he only examined Indo-Pacific material while Nel-
son included West African and New World material.
Evidence from eggs and larvae
There are two major problems with using characters of eggs
and larvae to criticize classifications based on adult characters.
First, the planktonic stages of fishes are exposed to different
selective pressures than the adults so they may show patterns
of specializations for planktonic life which are not congruent
with the distribution of adult character states. Second, relatively
few genera of clupeiform fishes have had the eggs or larvae
described for even one species in the genus. The first problem
must be dealt with the same as any character complex in a group
with more than one character complex. More knowledge of the
ecology of the larvae in the sea would indentify species with
different funtional requirements for their larvae. The second
problem may be resolved by using the available evidence in a
parsimonious fashion.
Eggs and young larvae are similar within genera. Seven species
of Sardine/la (Table 25) all have moderately sized clupeid-type
eggs with a wide perivitelline space and a single oil globule. The
egg described by Takita (1966) and Chang et al. (1981) as that
oi Harengida ziinasi is similar. Wongratana ( 1 980) places zunasi
in Sardinclla.
Within subfamilies there is little apparent consistency in egg
morphology among genera. Etruineus has no oil droplet but
Dussumieria does. Brevoortia has eggs 1.3 mm or larger with a
single oil globule; HHsa kelee has 1.00-1.07 mm eggs with sev-
eral small oil droplets. Clupea has demersal adhesive eggs while
Sprattus has pelagic eggs with a small perivitelline space. The
Indo-Pacific pristigasterin species of Ilisha have large eggs with
adhesive coatings and a single large oil globule but Opislhopterus
tardoore and the eastern Pacific O. dovii have small eggs with
small perivitelline spaces and no oil droplets.
The functional significance of egg characters is unknown. Sep-
arate lineages within the group which have radiated into several
habitats could show parallel adaptations such as oil droplets for
buoyancy or nutrition, adhesive coating for retention nearshore
or demersally. and egg size as a trade-off between broadcasting
and parental investment. Alternatively, different types of eggs
within taxonomic categories could also support splitting the
category. The anchovy genus Stolephorus contains species with
eggs which range from oval with no oil globule to varying degrees
of eccentricity with an oil droplet, to unusually shaped eggs with
knobs on one end (Delsman, 1931). Nelson (1983) separated
Stolephorus into two groups, a Stolephorus group with 1 3 species
and an Encrasicholina (new usage) group of 5 species which he
considered more closely related to New World anchovies than
to the 1 3 Stolephorus species. The three Encrasicholina species
whose eggs are known have an oval egg without a knob. One
of the three, E. hetcrolobus. was reported by Delsman (1931)
to have a small oil droplet and to be relatively more abundant
near shore than Stolephorus zolingeri. The other two, E. pur-
purcus and E. punctifer (^buccanceri, Strasburg, 1960; =zolin-
geri. Delsman, 1931), occur in Hawaii and neither has an egg
with an oil droplet. New World anchovies don't have eggs with
knobs or oil droplets; therefore, the evidence from eggs supports
Nelson's revision and in addition provides some basis for zoo-
geographic speculation.
Whether the pristigasterins should be given equal rank with
the clupeids and engraulids cannot be answered with the avail-
able ontogenetic information. There are two very different egg
types in the group, small with small perivitelline space and large
with gelatinous coating, both of which could be considered spe-
cializations. Etrumeus. Jenkmsia. Spratelloides, Clupea. Sprat-
tus, and Potamalosa were linked based on a foramen in the
fourth epibranchial (Nelson, 1970). Eggs of Spratelloides and
Clupea are both demersal and adhesive. The planktonic eggs of
Etrumeus and Sprattus both have narrow perivitelline spaces
and lack oil globules. Eggs of Potamalosa and Jenkinsia are
unknown. Jenkinsia is related to Spratelloides and has demersal
larvae (Powles, 1977) so it may have demersal eggs. The de-
velopmental osteology of these genera could be studied to de-
termine if the shared foramen is phylogenetically homologous.
The egg of Anodontostoma, Dorosominae, is similar to eggs of
the Alosinae in that it has multiple small oil droplets. Otherwise
both the Alosinae and Dorosomatinae contain species with de-
mersal adhesive eggs and species with buoyant planktonic eggs.
Other suggestions of Nelson (1970) that Sardinclla. Opistho-
nema. Harengula. and Herklotsichthys should be placed with
the Dorosomatinae and Sardina and Sardinops with the Alo-
sinae and then that the Alosinae and Dorosomatinae should be
combined leaving just Clupeinae and Dorosomatinae cannot be
critically evaluated with existing ontogenetic data. These hy-
potheses could be tested by comparing the osteological devel-
opment of the characters used by Nelson, augmented by other
early life history characters.
Relationships of the Clupeiformes
Greenwood et al., (1966) placed the Clupeomorpha and Elo-
pomorpha together in their Division One but gave serious con-
sideration to the possibility that the Clupeomorpha should be
recognized as a separate division. Using information on the gut
and lower jaw. Nelson (1973) proposed that the Clupeomorpha
were distinct from the Elopomorpha but perhaps related to the
non-osteoglossomorph teleosts. Gosline (1980) concluded that
the clupeiform fishes should be grouped with the elopiform, the
salmoniform, gonorynchiform, and ostariophysine fishes; sep-
arated on one side from the osteoglossiform fishes and from the
iniomous— acanthopterygian teleosts on the other. His conclu-
sions were based on five morphological character complexes:
the caudal skeleton, the swim bladder-ear connection, the post-
cleithrum, the structures associated with pectoral fin movement,
and the various types of premaxillary movements and jaw pro-
trusion (Gosline, 1980).
Gosline (1980) considered the elopomorphs to be an early
offshoot from a basal lower teleostean group. He considered the
gonorynchiforms and ostariophysines to be more closely related
to each other than to the clupeiforms. A clupeiform— osteo-
glossiform link has also been mentioned (Greenwood, 1973). J.
S. Nelson (1976), who put the superorders Clupeomorpha (Clu-
peiformes) and Elopormorpha (Elopiformes, Albuliformes, An-
guilliformes) into Division Taeniopaedia, slated succinctly that
"the relation of superorders recognized here is poorly known
and they are essentially "loose ends." " Lauder and Liem (1983)
provisionally follow Nelson (1970) for most groups within the
Clupeomorpha but represent the interrelationships of clupeoid
lineages as an unresolved polychotomy. Lauder and Liem (1983)
124
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 25. Sources of Early Life History Information for Clupeiformes. Reviews and readily available works with superior illustrations
are cited rather than original descriptions in some cases.
Genus species
Eggs
Lar-
vae
Ju-
ven- Mor- Mens-
lies phology tics
Pig-
menta- Oste-
Fins lion ology
Keys
or com-
pan-
sons Wild-
Fe- Spawn- Spawn- with Reared caught
cun- ing ing others speci- speei-
dity region season species mens mens
References
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Chirocentnis dorab X X
Chirocentrus nudus X X
Sardinella zunasi X X
Sardinella jussieui XXX
Sardinella aurila XXX
Sardinella albella X X
Sardinella fimbriata X X
Sardinella brachysoma X X
Sardinella brasiliensis X X
Sardinella longiceps X X
Sardinella maderensis X
Sardinella rouxi X
Clupea harengus XXX
Clupea pallasi XXX
Clupea bentincki X X
Spratlus sprattus XXX
Sprattus antipodurn X
Elrumeus teres X X X X
Elrumeus whiteheadi XX XX
Dussumieria sp. XX XX
Spratelloides delicatulus X X X X X
Jenkinsia lamprolaenia X X X X
Konosirus punctatus XX XX
Anodontostoma chacunda X X X X X
Dorosoma pelenense X X X X X
Amblygaster leiogasler XX XX
Amblygaster sirm X X
Escualosa thoracata XX X
Opisthonema lihenate X
Opisthonema oglinum X X X X X
Harengula jaguana X X X X X
Harengula peruana X
Sardinops sagax caerulea X X
Sardinops sagax musica X X
Sardinops melanosticta XX X
Sardinops ocellata X X X X X
Sardina pilchardus X X X X X
Lile stolifera X
Dorosoma cepedianum X X X X X
Hilsakelee XX XX
Tenualosa itisha XX XX
Alosa sapidissima X X X X X
Alosa pseudoharengus X X X X X
X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X
X X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
XXX
XXX
X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X
XXX
XXX
X
X X
X
X X
X X
X
X
X
X
Delsman, 1923, 1930b
Delsman, 1923, 1930b
Takita, 1966; Chang
et al., 1981
Bensam, 1970
Jones et al., 1978;
Houde and Fore, 1973
Delsman, 1933b
Delsman, 1926
Delsman, 1926
Matsuura. 1975
Nair, 1960
Conand, 1978; Conand
and Fagetti, 1971
Conand, 1978
Jones et al.. 1978;
Fahay, 1983
Wang, 1981
Orcllana and Balbontin,
1983
Saville, 1964
Russell, 1976; Robert-
son, 1975a
Mito, 1961a
Brownell, 1979; OToole
and King, 1974
Delsman, 1925
Uchidaet al., 1958;
Miller et al., 1979
Powles, 1977
Mito, 1961a
Delsman, 1933a
Shelton and Stephens,
1980; Jones etal.,
1978
Delsman, 1926b
John, 1951a
Delsman, 1926c, 1934a
Peterson, 1956
Richards et al.. 1974;
Jones et al., 1978
Houde etal.. 1974;
Gorbunova and
Zvyagina 1975;
Houde and Fore, 1973
Peterson, 1956
Ahlstrom, 1943; Miller,
1952
Santander and de Castillo,
1977; Orellanaand
Balbontin, 1983
Mito, 1961a
Brownell, 1979; Louw
and OToole, 1977
SaviUe, 1964; Russell,
1976
Peterson, 1956
Shelton and Stephens,
1980; Jones etal.,
1978; Cooper, 1978
Rao, 1973
Kulkami, 1950
Bainbridge, 1962;
Jones etal., 1978
Jones etal., 1978;
Chambers et al., 1976
McGOWAN AND BERRY: CLUPEIFORMES
125
Table 25. Continued.
Genus species
Eggs
Ur-
vae
ven- Mor- Mens-
iles phology tics Fins
Pig- Fe-
menta- Oste- cun-
tion ology dity
Keys
or corn-
pan -
sons Wild-
Spawn- Spawn- with Reared caught
ing ing others speci- speci-
region season species mens mens
References
Alosa mediae ns
Alosa aestivalis
Caspialosa sp.
Elhmalosa fimbriala
Brevoortia aurea
Brevoortia patronus
Ethmidium macutata
Gilchristella aesluanus
Laevisculella dekimpei
PeUonula vorax
Ilisha elongata
Ilisha melasloma
IHsha afncana
Ilisha furthi
Neoopislhopterus tropicus
Opisthopterus tardoore
Opisthopterus do\i
Opisthopterus equatorialis
Odontognathus panamensis
Anchoa ischana
Anchoa panamensis
A nchoa curta
Anchoa tucida
Anchoa naso
Anchoa exigua
A nchoa arenicola
Anchoa marinii
Anchoa hepsetus
Anchoa mitchilli
Anchovia macrolepidota
Engraulis japo nicus
Engraulis eur\'slole
Engraulis anchoita
Engraulis inordax
Engraulis encrasicolus
Engraulis ringens
Slolephorus purpureus
Stolephorus buccaneeri
Stolephorus heterolobus
Slolephorus tri
Thryssa hamiltonii
Thry'ssa sp.
Lycengraulis poeyi
Lycengraulis gross idens
Celengraulis mysticetus
Setipinna melanochir
Selipmna taty
Setipinna phasa
Heterothrissa breviceps
Coilia sp.
Coilia sp.
X
X
X
X
X
X
X
X
X
X
X
X
X
X
XXX
XXX
X X
X
X
X
X
XXX
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X X X X X X
X X X X X X
X X
X X
X X
X X
X
X X
X X
XX X
XX X
XX X
XX X
XXX X
XX X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X Jones etal., 1978;
Chambers et al., 1976
X X Jones etal., 1978
Pertseva, 1936
X X Bainbndge, 1961
X Conand 1978; de
Ciechomski. 1968
X Houde and Fore, 1973
X Orellana and Balbontin,
1983
X Brownell, 1979
X Conand, 1978
Bainbndge. 1962;
Conand, 1978
X X Delsman, 1930a; Uchida
etal.. 1958
X Delsman, 1930a
X Dessier, 1969
Peterson. 1956
Peterson, 1956
X Bensam, 1967
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
Peterson, 1956
de Ciechomski, 1968
Jones etal., 1978
Jones etal.. 1978
Peterson, 1956
Mito, 1961a; Brownell,
1979; Russell, 1976
Jones etal., 1978
de Ciechomski, 1965
Bolin. 1936a; Ahlstrom,
1965; Ahlstrom,
unpublished
X D'Ancona, 1931a; Saville,
1964; Marchal, 1966
X X Orellana and Balbontin,
1983; Fischer, 1958b;
Einarsson and Rojas
de Mendiola, 1963
X Miller etal., 1979
X Delsman, 1931; John,
1951a
X Delsman, 1931
X Delsman, 1931; John,
1951a
X Delsman, 1929a
X John, 1951a
X Peterson, 1956
X Phonlor, 1978
X Simpson, 1959
X Delsman, 1932a
X Delsman, 1932
X Jones and Menon, 1950
X Delsman, 1932a
X Takita, 1967
X Delsman, 1932b
X
X
X
X
X
X
X
X
X
X
X
X
X
X
126
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
place the clupeomorpha nearer to the next most advanced clade,
the Euteleostei, than to the next least advanced clade, the Elo-
pomorpha.
Evidence from eggs and larvae
Relevant ontogenetic evidence concerning the relationships
of the Clupeiformes is meager. Elopiform eggs are unknown.
Anguilliform eggs resemble clupeid eggs in having perivitelline
spaces, segmented yolks, and may have oil droplets. Eel eggs
can be much larger than herring eggs: 5.5 mm diameter in A/m-
raena. 2.43 mm in an anguillid (Ahlstrom and Moser, 1980).
Osteoglossomorphs have pelagic or demersal eggs which may
be 0.5-4.0 mm in diameter, may be dark blue, and may have
a very wide perivitelline space as in Hiodon (Breder and Rosen,
1966). The coincidence of demersal adhesive eggs in both the
osteoglossomorphs and the Dorosomatinae is extremely un-
likely to be a shared derived character from a common ancestor.
Clupeid and anguillid eggs are considered unspecialized relative
to eggs of the higher teleosts (Ahlstrom and Moser, 1980). Very
little else may be said. Perhaps electron microscopy will reveal
patterns of chorion sculpturing which will be informative.
The larvae of Clupeiformes are unspecialized and undergo a
fairly uneventful metamorphosis. The migration of the dorsal
fin during transformation also occurs in the elopiforms. The
larva of Chanos. a primitive gonorynchiform (Fink and Fink,
1981), superficially resembles clupeids or engraulids but appar-
ently does not have the same migration of the dorsal fin (Rich-
ards, this volume).
If the Elopomorpha and the Clupeomorpha share a common
ancestor it is possible that the Clupeomorpha retained the un-
specialized, rapidly developing larvae while the adults evolved
towards a specialized schooling planktivore body plan. The lep-
tocephalus found in the elopiforms, albuliforms, and anguilli-
forms could have evolved for dispersal or to reduce predation
or to take advantage of larval drift the way Angntlla does in the
North Atlantic and the way herring do in the North Atlantic
with their circuit of migration (Cushing, 1977). The leptoceph-
alus could have arisen in the common ancestor of anguilliforms
and elopiforms or in parallel, in response to the same selective
influence, after the adult eels had begun their divergence from
the still unspecialized elopiform fishes. The leptocephalus is
considered a specialized character by Forey (1973a), who sug-
gested that it arose before the elopid-albulid dichotomy. Trans-
forming elopoid leptocephali resemble transforming clupeiform
larvae (A/e^a/ops— Harrington, 1958: Plate 1; f/ops— Sato and
Yasuda, 1980: Fig. 1; ,4//)j//a-Hildebrand, 1963b: Fig. 23).
The egg and larval evidence thus is consistent with a rela-
tionship between the Elopomorpha and the Clupeomorpha based
on primitive characters but is not helpful in aligning this Di-
vision (J. S. Nelson's usage, 1976) closer to any other.
Summary and recommendations
The eggs and early larval stages of the Clupeiformes provide
many taxonomic characters with potential value for testing phy-
logenetic hypotheses. Most of the discrete characters, such as
number of oil globules, have more than two states and the
continuous characters, such as degree of egg eccentricity, have
at least a moderate range of values. Although the fraction of
species whose eggs and larvae have been described is low and
the descriptions are uneven in quality and not distributed uni-
formly among taxa, egg and larval characters appear consistent
within genera. Within nominal subfamilies they are not consis-
tent, but the subfamilies show parallel trends in adult characters
and, in addition, the distribution of genera in higher taxa is not
yet agreed upon by all workers.
Most descriptions of clupeiform larvae have been to enable
identification of regional species. Differences between larvae
usually involve subtle features of pigmentation or morphome-
try, or counts of meristic characters which converge with the
meristics of the adult. Phylogenetically significant characters
such as ephemeral dentition, osteological development, and the
comparative ontogeny of characters used in the taxonomy of
the adults are rarely mentioned.
Future descriptions of eggs and larvae should address system-
atic characters as well as those needed for identification. Eggs
and larvae of many species should be redescribed to give com-
plete series through metamorphosis. Ontogenetic characters
should be used in revisions of the group. Classifications of the
Clupeiformes which are based on just a few characters should
be tested by comparing the ontogeny of those characters because
there are many apparently parallel trends in the group. Addi-
tional studies of the physiology and ecology of the eggs and
larvae should be done to determine the functional significance
of observed characters. It would also be useful to perform quan-
titative phenetic and cladistic analyses now of the Clupeiformes
for those regions or taxa for which information is already fairly
complete.
National Marine Fisheries Service, Southeast Fisheries
Center, 75 Virginia Beach Drive, Miami, Florida 33149.
Ostariophysi: Development and Relationships
L. A. FUIMAN
OSTARIOPHYSI, as regarded here, include all fishes whose 3 orders, about 55 families, and more than 5,000 species, there-
four or five anteriormost vertebrae are modified to form by accounting for over 70% of the world's freshwater fish species,
an otophysic connection, the Weberian apparatus (Rosen and Oslariophysans occupy most freshwater habitats worldwide, from
Greenwood, 1970). These primarily freshwater fishes comprise torrential Himalayan streams to still tropical lakes, as well as
FUIMAN: OSTARIOPHYSI
127
Fig. 6 1 . Egg of Clenolucius hujela ( 1 8 hours poslfertilization) show-
ing the membranous pedestal by which the egg attaches to plants. Pho-
tograph by H.-J. Franke.
coastal marine waters (the latter by a few characids, cyprinids,
and aspredinids, as well as all ariid and plotosid catfishes). The
presence of a Webenan apparatus has overshadowed the suite
of remaining diagnostic characters for the group which includes
an axe-shaped endochondral portion of the metapterygoid, an-
teriorly bifurcate pelvic girdle, second hypural fused to the com-
pound terminal centrum, and elongate olfactory tracts (all de-
tailed by Fink and Fink, 1981). Additional characters include
a pheromone-mediated alarm reaction and homy dermal pro-
jections called unculi (Roberts, 1982b).
According to the classification of Fink and Fink (1981), the
orders of Ostariophysi (their Otophysi) are: Cypriniformes,
Characi formes, and Siluriformes (the latter including Siluroidei
and Gymnotoidei). Cypriniforms (with over 1,800 species in 5
families) uniquely share peculiarities of the following: kineth-
moid bone, palatine-mesopterygoid articulation, fifth cerato-
branchial, and lateral process of the second vertebral centrum.
They lack jaw teeth and an adipose fin. They are found in North
America, Eurasia, and Africa. Characiforms (comprising at least
1,000 species in 14 families) are characterized by multicuspid
teeth, a prootic foramen, dorsomedial opening in the posttem-
poral fossa, enlarged lagenar capsule, and a gap between the
compound terminal centrum and hypural 1. They occur in Af-
rica, South America, and southernmost North America. Silu-
roids (with about 2,000 species in 3 1 families) are distributed
nearly worldwide. Although quite diverse morphologically, they
commonly lack scales and several bones (including the sym-
plectic, subopercle, and separate parietals). They show consid-
erable fusion of portions of the first five vertebrae and pectoral
and dorsal fin rays. The electrogenic gymnotoids are character-
ized by an extremely long anal fin and substantial reductions or
losses, such as the loss of dorsal and pelvic fins, and palatine
and ectopterygoid bones. They are confined to South America
and southernmost North America.
Development
Knowledge of the early life history stages of ostariophysans
is rather spotty and concentrated on fishes from a few geographic
regions. Major descriptive works cover portions of the Soviet
Union (Kryzhanovskii, 1949; Kryzhanovskii et al., 1951; Kob-
litskaia, 1981), Japan (Okada, 1960; Nakamura, 1969), and the
United States (Jones et al., 1978; Snyder, 1981; Auer, 1982;
Fuiman et al., 1983). Most of these works concentrate on cy-
priniforms. Additional descriptive data are available as indi-
vidual papers on Indian major carps (Cyprinidae) and Indian
siluroids (reviewed by Jhingran, 1975). African and South
American ostariophysan eggs and larvae remain little known.
Of the six families of cypriniforms, nothing is known of the
eggs and larvae of the families with fewest species, Gyrinocheili-
dae and Psilorhynchidae. Catostomids are known well. Cypri-
nids, cobitids, and homalopterids are known to a lesser degree.
Scattered notes are available for nine characiform families but
only a few descriptions of ontogeny exist. Brief descriptions of
larvae of representatives from seven families of siluroids are
available, and notes on eight additional families exist. Photo-
graphs of larvae of two gymnotoids. Eigenmannia virescens anA
Aptewnotus leptorhynchus are published (Kirschbaum and
Westby, 1975; Kirschbaum and Denizot, 1975; Kirschbaum,
1984) but without morphological descriptions. Most informa-
tion on ostariophysan larvae deals with external morphology.
Osteological studies are few (Bertmar, 1959; Hoedeman, 1960a-
d).
Eggs
Ostariophysan eggs vary considerably in their morphology
and the habitat they occupy. Most are spherical, demersal, 1 to
5 mm in diameter, with pale yellow, somewhat granular yolk
Table 26. Larval Characters of Major Groups of Ostariophysans.
Cypnniformes
Characiformes
Siluroidei
Gymnotoidei
Size at hatching (mm XL)
Yolk-sac shape
Gap between yolk sac and anus
Barbels:
Presence
Timing of development
Size at finfold absorption (mm TL)
2-10
pyriform or tubular
absent
present or absent
late or early
15-25
2-5
elliptical
present
absent
10-20
3-8
elliptical
present
present
early
11-23
elliptical
absent
absent
15
128
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
A
FUIMAN: OSTARIOPHYSI
129
Ssur
Fig. 62. Representative cypriniform larvae. (A-C) Cyprinidae: (A) Tribolodon hakonensis (UMMZ 212151) 9.2 mm TL; (B) Semotilus
alromaculatus 8.6 mm TL; (C) Barbiis { = Capoela) tilteya (UMMZ 212148) 6.0 mm TL; (D, E) Cobitidae; (D) Misgurnus fossiHs 6.9 mm TL,
after Kryzhanovskii (1949); (E) Acanthophthalmus cf kuhni 4.0 mm TL (specimen from S. S. Boggs).
lacking oil globules. Eggs may be strongly adhesive (e.g., Cy-
priniformes: Nemacheilus [=Barbatula] torn [Kobayasi and
Moriyana, 1957]; Characiformes: Gymnocon'mhus tenictzi [pers.
obs.]; Siluriformes: Loricana calaphracta [pers. obs.]), nonad-
hesive (e.g., Cypriniformes: Clenopharyngodon idclla [Inaba et
al., 1957]; Siluriformes: Tandanm landanus [Lake. 1967]), or
weakly adhesive (e.g., Cypriniformes: Catoslomus commersoni
[pers. obs.]; Characiformes: Scrrasalmm nattercn [pers. obs.];
Siluriformes: Baganus hagarius [David, 1961]). Adhesive fila-
ments or other apparent modifications of the egg surface are
almost entirely unknown.
Representatives of outgroups (Gonoi^nchiformes, Clupeo-
morpha, "Salmoniformes," and Osteoglossomorpha) share the
spherical egg with yellow, granular or segmented yolk. Their
eggs are pelagic or demersal, usually 1 .0 to 1.3 mm in diameter.
adhesive (in Osmerus) or nonadhesive (in Chanos. Alosa. and
Hiodon). without oil globules (Chanos) or with one to several
(in Alosa and Osmerus).
Exceptions to this characterization of ostariophysan eggs exist.
Among cypriniforms, the cyprinid subfamily Acheilognathinae
(Gosline, 1978) exhibits elliptical to pyriform eggs which are
deposited in the mantle cavity of a bivalve mollusc (Kryzhan-
ovskii et al., 1951; Nakamura, 1969; Makeeva, 1976). Their
irregular shape may be the important mechanism preventing
the eggs from being expelled. Some cyprinid eggs are pelagic
(e.g., Hypophthalmichthys molitrix [Nakamura, 1969; Koblit-
skaia, 1981]) and have a larger diameter (ca. 5 to 6 mm) due
to the considerable perivitelline space. Only one ostariophysan,
the cypriniform Cobitis biwae, was reported to have 12 to 13
small oil globules in the yolk (Okada and Seiishi, 1938; Okada,
Fig. 63. Representative cypriniform larvae (continued). Catostomidae: Hypentetium etowanum (upper) 13.1 mm and (lower) 15.0 mm TL.
130
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
/JS-^y
Fig. 64. Representative characiform larvae. Serrasalmidae: Serrasalmus nattereri (UMMZ 211677) 8.2 mm TL (upper). Characidae: Hy-
phessobrycon cf. callistus (UMMZ 21 1676) 6.6 mm TL (lower).
1960), but this is in doubt (N. Komada, pers. comm.) and has
not since been confirmed.
Characiform eggs are poorly known; most information is from
the aquarium hobby literature. Known characid (sensu Gery,
1977) eggs are small (0.8 to 1.2 mm). However other families
have eggs between 2 and 4 mm (e.g., Alestidae, Anostomidae,
Curimatidae, Hepsetidae, Serrasalmidae). Apparently most
species have eggs that adhere to plants. Franke (1981) described
adhesive threads (gallertigen Klebfdden) on the surface of the
egg of Ctenolucius hujeta (Ctenoluciidae) and noted that this
was the mechanism by which they attached to plants. My ex-
amination of eggs supplied by Dr. Franke found the adhesive
structure to be a membranous pedestal rather than adhesive
threads (Fig. 6 1 ). This is the only known chorionic modification
of ostariophysans.
Most siluroids have demersal, medium sized eggs ( 1 to 4 mm).
Some are tended by one or both parents [e.g., Clarias batrachus
(Mookerjee, 1946; Mookerjee and Mazumdar, 1950), Ictalurus
punctatus (Tin, 1982c)]; others are not given parental care [e.g.,
Clarias gariepinus (HoW, 1968; Bruton, 1979), Pangasius sutchi
(Varikul and Boonsom, 1969)]. The eggs are typically spherical;
however, Clarias eggs are often slightly elliptical (Mookerjee,
1946; Greenwood, 1955; Bruton, 1979). Some callichthyids de-
posit small eggs (ca. 1.0 mm) in a foam nest on the surface of
still waters (Kryzhanovskii, 1949). Parents in several families
carry their eggs. Some loricariids (e.g., Loricaria spp.) carry a
mass of eggs by means of fleshy appendages of the lower lip.
Aspredo laevis eggs apparently are attached by vascularized stalks
to the venter of the female (Wyman, 1859). Finally, ariids are
oral incubators with perhaps the largest eggs of all oviparous
teleosts (10 to 25 mm) (Chidambaram, 1942; Gudger, 1912,
1916, 1918; and other authors). Although yolk is usually yellow
to slightly orange or brown, several species have unmistakably
green yolk [e.g., Bagarius bagarius (David, 1961), Clarias ba-
trachus (Mookerjee, 1946; Mookerjee and Mazumdar, 1950),
Heteropneustes fossilis (Pal and Khan, 1969), Loricariichthys
sp. (Taylor, 1983), Phractura ansorgei (Foersch, 1966)]. At least
one siluroid, the silurid Ompok bimaculatus, has reddish brown
yolk (Chaudhuri, 1962). A few species have a jelly-like coat
surrounding the chorion [e.g., Bagarius bagarius (David, 1961),
Parasilurus asotus (Kryzhanovskii et al., 1951), Phractura an-
sorgei (Foersch, 1966), Trachycorystes insignis (Burgess, 1982)].
Larvae
Most ostariophysans hatch in an altricial state at about the
time when pectoral buds form, but before the head becomes
free from the yolk sac and retinal pigment develops, although
there is variability in the exact stage. The yolk sac is usually
large and cumbersome, enforcing a stationary existence during
the first days, either on the substrate (most commonly) or at-
tached to plants by means of a cephalic adhesive mechanism
(found in most characiforms and a few cyprinids, but structur-
ally diflTerent in these groups). Caudal fin rays diflierentiate first,
followed by nearly simultaneous formation of dorsal and anal
fin rays. Pectoral and pelvic fin rays develop near the end of the
larval period. The gonorynchiform Chanos hatches at about the
same stage of development as ostariophysans, but Atosa and
Osmerus hatch somewhat later (i.e., pectoral buds and retinal
pigment are clearly developed). These outgroups generally have
pelagic larvae at hatching. Fin rays in Chanos develop in the
FUIMAN: OSTARIOPHYSI
131
same order as described above, but the sequence differs for Alosa
and again for Osmerus.
Within Ostariophysi, cypriniform larvae (Figs. 62, 63) are
largest at hatching (Table 26), the largest sizes represented most-
ly by catostomids. The pyriform yolk sac extends from below
the head posteriorly to the anus (Fig. 62a). Barbels, when pres-
ent, develop very late in Cyprinidae but early in Cobitoidea
(sensit Sawada, 1982). Cyprinids display considerable variation
in the elaboration of the larval circulatory system. Temporary
networks of blood vessels invade portions of the finfolds and
the surface of the yolk sac in a variety of patterns to form the
larval respiratory system (Kryzhanovskii, 1947). Cobitoideans
usually have greatly expanded finfolds, especially those of the
pectoral buds. Pronounced external gill filaments are known in
the cobitine genera Coto/5 (Kryzhanovskii, 1949;Okada, 1960;
Sterba, 1962), Lepidocephaliis (Bhimachar and David. 1945),
and A/;5^r«wi (Kryzhanovskii, 1949; Okada, 1960), but not in
the non-cobitine cobitoidean genera Botta. Lefua, or Nemach-
eilus, nor in other ostariophysans. Cyprinids with cephalic ad-
hesive glands include: Ahramis brama (Penaz and Gajdusek,
1979); Brachydanio rerio (Frank. 1978); Cypri niis carpio (Hoda
and Tsukahara, 1971); Danio malabancus (Jones, 1938); and
Notemigonuscrysoleucas (Snyder tXa\., \911\ Loosetal., 1979).
In characiforms, the yolk sac is short and rounded, not ex-
tending to the anus posteriorly (Fig. 64). Most known characids
(sensii stricto) and a hepsetid (Bertmar, 1959; Budgett, 1902.
1 903), erythrinid (de Azevedo and Gomes, 1 942), and curimatid
(de Azevedo et al., 1938) have a temporary larval cephalic ad-
hesive organ (more distinct than the apparent glandular mech-
anism in cyprinids). Those without such an organ mclude: Ser-
rasalmus nattereri (pers. obs.), Metynnis maciilatiis (Azuma,
1982), and Brycinus longipinnis (Frank, 1972). The adipose fin
appears to develop de novo toward the end of the larval period,
not as a remnant of the median finfold. However, the small size
of the adipose fin and lack of specimens, photographs, illustra-
tions, and descnptions of late larval characiforms prevents ver-
ification of this inference.
Although few species are known as larvae, Siluroidei may
contain the greatest diversity of larval characters among Ostar-
iophysi (Fig. 65). Most siluroids hatch as altricial larvae with a
physiognomy similar to that of characiforms. Ictalurids are more
precocial and lack a postlarval (sensu Hubbs, 1 943) phase. Ariids
(Gudger, 1918; Ward, 1957) and some loricariids (Lopez and
Machado, 1975; Machado and Lopez, 1975) hatch in a highly
precocial state, resembling the adult in many aspects of external
morphology but retaining a large yolk sac (Fig. 65C). In most
families, barbels are usually present at hatching or soon there-
after (Fig. 65a). Cephalic adhesive organs are usually absent,
but at least one loricariid (Ancistrus sp.) possesses these (Franke.
1979). Clarias gariepinus (=C. mossambicus) and Ompok bi-
maculatus have an adhesive organ on the venter of the yolk sac
(Greenwood, 1955, 1956; Chaudhuri, 1962; Holl, 1968;Bruton,
1979). The adipose fin is clearly a remnant of the median fin-
fold, as in "'salmoniforms." Larvae of a single gymnotoid, Ei-
genmannia virescens. are known (Fig. 65D, E; Table 26; Kirsch-
baum and Balon, in prep.).
Relationships
The Ostariophysi are thought to be the sister group of the
Gonorynchiformes (Greenwood et al., 1966; Rosen and Green-
wood, 1970; Gosline, 1971; Fink and Fink, 1981). The next
closest relatives are Clupeiformes (Gosline, 1971) or "Salmon-
iformes" (Greenwood et al., 1966; Fink and Weitzman, 1982).
All concepts of Ostariophysi (those with a Weberian appa-
ratus) recognize four major groupings, "cyprinoids," "chara-
coids," "gymnotoids," and "siluroids." The traditional view of
relationships holds that "characoids" are the ancestral stock,
giving rise to the remaining lineages, with "gymnotoids" being
modified "characoids," and "cyprinoids" being the closest rel-
atives of the "characoids" plus "gymnotoids." Fink and Fink
(1981) gave a detailed history of the classification schemes for
the Ostariophysi and their relatives as an introduction to their
work on the subject, which is the only attempt to reconstruct
the phylogeny on the basis of a large set of data ( 1 27 characters).
Their proposed cladistic phylogeny differs significantly from the
traditional one by aligning "gymnotoids" with "siluroids" as
the Siluriformes (Fig. 66).
Developmental characters in systematics
Few attempts have been made to apply developmental char-
acters to the systematics of ostariophysans. Kryzhanovskii (1947)
grouped cyprinids into four subfamilies according to details of
the larval respiratory system. He also included characters re-
lating to reproductive guild (later elaborated in Kryzhanovskii,
1948), original (ontogenetically) position of the mouth, and rel-
ative size of the pectoral buds. He supported these subfamilial
designations with experimental results on the morphology and
viability of larvae produced by artificial hybridizations within
and among the proposed subfamilies.
Nakamura (1969) dealt with the cyprinids of Japan. In his
English summary, he stated that currently proposed closely re-
lated forms (meaning genera, species, and subspecies) have sim-
ilar life history characteristics. He noted a few exceptions, such
as similar (as adults) species oi Moroco whose early larvae differ
morphologically and ecologically. In contrast, he noted that the
eggs and early larvae of Ctenopharyngodon idella and Hypoph-
thalmichthys molitri.x were very similar although the species
were placed in different subfamilies. He used differences in egg
and larval morphology to support the previously uncertain sep-
aration of the genera Squalidus and Gnathopogon.
In a similar survey. Loos and Fuiman (1978) attempted to
characterize the subgenera of the New World cyprinid genus
Notropis in terms of their egg and larval morphology. However,
they found substantial variability within the established sub-
genera and were unable to characterize them precisely.
Each of these attempts to apply developmental characters to
systematics was concerned only with establishing group mem-
bership and not with determining relationships among the groups.
Further, none of the work was based on a large data set nor was
it approached in a rigorous manner. The difficulties encountered
by Nakamura ( 1 969), and especially by Loos and Fuiman (1978),
probably were due to the apparently convergent ecomorpho-
types expressed by unrelated taxa. The low taxonomic level
investigated, combined with the morphological similarity im-
plied by von Baer's law, probably accounted for much of the
remaining difficulty in detecting consistent differences among
taxa.
Fink and Fink's (1981) classification is based largely on os-
teological characters. The great size and diversity of Ostario-
physi make a detailed study of developmental osteology and
concomitant investigations of bone homologies impractical at
this time. Yet, available information permits a preliminary eval-
132
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
D
FUIMAN: OSTARIOPHYSI
133
Fig. 65. Representative siluriform larvae. (A-B) Clariidae: Clanas gariepinus (British Museum of Natural History, uncataloged) (A) 6.6 mm
and (B) 8.4 mm TL; (C) Loricariidae: Ancistrus spinosus (UMMZ 212152) 8.3 mm TL; (D-E) Rhamphichthyidae; Eigenmannia virescens (D)
5.0 and (E) 8.1 mm TL.
uation of relationships based on developmental characters. The
following analysis attempts to evaluate the contribution of se-
lected developmental characters to ostariophysan systematics
by constructing an independent assessment of phylogeny based
on developmental characters. That the assessment should be
independent was attested by Moser and Ahlstrom (1974): "we
are increasingly impressed with the functional independence of
larval and adult characters. It is apparent that the world of the
larvae and the world of the adults are two quite separate evo-
lutionary theaters."
Representative ontogenetic series of all families of ostario-
physans are nearly impossible to obtain because of the large size
and wide geographic distribution of the group and the dearth
of ichthyologists studying larvae. Consequently, the analyses
employed here were based on specimens generated from labo-
ratory breeding experiments, wild-caught material, and data
published in apparently accurate accounts of ontogeny. Species
used in the analyses included four outgroups to the Ostariophysi
(Gonorynchiformes, Clupeomorpha, "Salmoniformes," and
Osteoglossomorpha), all characiforms and siluriforms with suf-
ficient morphometric and developmental data for analysis, and
a sample of five species from the most primitive cypriniform
family, Cyprinidae. These cyprinid species possess different
combinations of larval characters (determined by their location
on a Wagner tree generated for 33 larval cyprinids [Fuiman,
1983a]). Although not used directly, incomplete data on ap-
proximately 85 additional non-cyprinid ostariophysans provid-
ed corroborative information.
Species included in the analysis of relationships and their
sources are listed below. Initials denote specimens borrowed
from, or information provided by: Florida State Board of Con-
servation (FSBC), University of Michigan Museum of Zoology
(UMMZ), or Frank Kirschbaum (FK).
OsTEOGLOSSiFORMEs: Hiodofi tergisus [Snyder and Douglas
(1978); Wallus (1981, pers. comm.)].
Salmoniformes: Osmerus mordax [Cooper ( 1 978); Tin ( 1 982b)].
Clupeiformes: Alosa pseudoharengus [Jones et al. (1978); Tin
(1982a)].
Gonorynchiformes: Chanos chanos [Chaudhuri et al. (1978);
Liaoet al. (1979); Miller et al. (1979)].
Cypriniformes: Cyprinidae— Cvpn>!Wicarp/o [UMMZ 21 1678;
Hoda and Tsukahara (1971); Nakamura (1969); Okada
(I960)]; Leiiciscus cephaliis [Cemy (1977); Kryzhanovskii
(1949); Penaz (1968); Prokes and Penaz (1980)]; Opsan-
ichthys unciroslris [Kryzhanovskii et al. (1951); Makeeva
and Ryabov (1973); Nakamura (1951, 1969)]; Parabramts
pekmensis [Institute of Hydrobiology (1976); Kryzhanov-
skii et al. (1951)]; Squalidus gracilis [Nakamura (1969)].
Characiformes: Alestidae— .-l/eirw haremose [Durand and
Loubens (1971 )]. Erythrinidae— //op/Zaj^ malabaricus [FSBC
8962, 8963, 9593; de Azevedo and Gomes (1942); Hensley
(1976); Moreira ( 1 920); von Ihering et al. ( 1 928)]. Charac-
idae— Hyphessobrycon cf. callistiis [UMMZ 21 1676], Ser-
r&sdAmiddie— Serrasalmus nattereri [UMMZ 21 1677; Azu-
ma(1975)].
Siluriformes: Siluroidei: Ba.gn6.aQ — Mystus seenghala [Saigal
and Motwani (1962)]; Rita rila [Karamchandani and Mot-
Cypriniformes
Characiformes
Siluroidei
Gymnotoidei
Fig. 66. Cladogram of ostariophysan relationships derived from adult characters by Fink and Fink (1981). Stem lengths imply no special
significance.
134
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Squalidus
CYPRINIFORMES
Ictalurus
21
Eigenmannia
*28
SILURIFORMES
CHARACIFORMES
Fig. 67. Wagner tree of ostariophysan phylogeny based on larval characters. Stem lengths are proportional to the number of character-state
changes on a given stem.
wani (1955)]. Clariidae— Ctor/a^ batrachus [UMMZ 1 86690.
209039; Devaraj et al. (1972); Mookerjee (1946); Mook-
erjee and Mazumdar (1950)]. Ictaluridae— /rta/Mr«5 neb-
ulosus [Armstrong ( 1 962); Tin ( 1 982c)]. Pangasiidae-Paw-
gasius sutchi [Varikul and Boonsom (1969)]. Sisoiidae—
Bagarius bagarius [David (1961)]. Gymnotoidei; Rham-
ph'ichVnyiddie— Eigenmannia virescens [FK, Kirschbaum
and Westby (1975)].
Phylogenetic methods
The phylogenetic reconstruction based on developmental
characters was generated by the cladistic Wagner tree method
(described by Kluge and Farris. 1969; Farris. 1970; Lundberg,
1972; and Jensen, 1981). Characters were chosen by virtue of
their availability in published accounts. Nearly all were recorded
as continuous measures, but individual modes with their neigh-
boring values and disjunct portions of distributions separated
Table 27. Ranges of Values for Coded Character States of 16 Ostariophysans. Character numbers correspond to those given in the
text. Primitive states are given in boldface type.
\*V\z\rz\(^\PT
Character slate
number
a
b
c
d
e
f
1
2.27-3.42
3.74-3.74
4.74-4.90
2
3.93-5.06
5.65-6.28
7.06-7.40
8.86-8.86
3
1.35-1.53
1.92-2.58
2.98-4.03
4
1.19-1.19
2.09-2.09
2.46-3.39
3.78-5.14
5
0.97-1.61
2.18-2.18
6
2.06-2.46
2.62-2.86
3.25-3.31
7
0.13-0.28
0.36-0.52
8
0.14-0.28
0.42-0.58
0.71-0.82
9
0.72-1.00
1.29-1.29
10
1.26-1.68
1.97-2.11
11
-1.22—0.94
-0.67—0.23
0.06-0.06
12
0.70-1.04
1.15-1.40
13
0.22-0.39
0.54-0.84
14
15.3-19.0
21.6-22.5
25.0-26.3
28.5-30.3
32.7-32.7
15
8.0-8.0
12.0-20.0
25.5-25.6
29.0-29.0
38.7-38.7 45.9-45.9
16
0.28-0.29
0.39-0.44
0.52-0.70
0.81-0.81
17
0.35-0.35
0.52-0.55
0.70-O.96
1.06-1.10
18
0.22-0.38
0.42-0.56
19
0.0-0.05
0.12-0.28
0.44-0.44
20
0.0-0.03
0.07-0.15
0.22-0.33
21
0-0
1-1
22
0-0
1-1
23
0-0
1-1
24
0-0
1-1
FUIMAN: OSTARIOPHYSI
135
Table 28. Character-State Changes on Stems Leading to
Hypothetical Ancestors (Nodes) and Terminal Taxa on the
Wagner Tree of Ostariophysi. Numbered character states correspond
to those given in Table 27 , Uniquely derived, unreversed character states
are given in boldface type. Reversed characters are noted by (r). Node
numbers correspond to those given in Fig. 67.
Node Characler state
1 8c, lib, 12b, 14b, 20b, 20c 24b
2 6b
3 14d, 15b, 18b
4 13a, 21a
5 18a(r), 20b(r)
6 16b
7 17b
8 14c, 20a(r), 23b
9 6b, 19a. 24a(r)
10 3b, lla(r), I2a(r). 14b
11 lla(r), 22b
12 6b, 6c
13 3b, 4c, 12b(r), 14a, 14b, 14c
14 3a, 4b, 18a(r), 20b(r)
15 10b, lib, 24a(r)
16 12b. 17b. 14b(r)
17 2b, 6b, 7a. 8b(r)
18 8a, 16b. 20b{r)
19 12b, 16a, 18a(r), 20a(r)
20 2a(r), 6a(r). lOa(r)
2 1 5b, 6c, 7b(r), 8b(r), 9b, 1 1 c, 1 5c, 17a, 1 7b. 1 9a
22 lOa(r), 15c, 15d
23 14b(r)
24 lb, 2c, I7d
25 6a(r), 1 la(r), 15c(r), 16d, 19c
26 Ic, 15e, 19a, 20b(r). 22a
27 2d, 8a, 14b(r), 14c(r), 17b. 17c(r). 18a(r)
28 4a, 4b, 6c, 10b, 15f
by measurable gaps were coded individually. Characters were
polarized by outgroup comparison (Table 27). The evolutional^
transformation series for each continuous, multiple state char-
acter was assumed to be linear (i.e.. with one or two adjacent
states for a given state). Consequently, a character coded with
n states had n - 1 different changes from one state to another,
disregarding the direction of change. These transitions were
termed "two-state factors." All two-state factors and their states
for each species were generated by the FACTOR computing
program (Estabrook et al., 1976). The output from this program
included an input file for the WAGNER 78 computing program
which was used to construct Wagner trees. The data deck was
resequenced and a new Wagner tree generated several times in
order to identify the shortest (most parsimonious) tree (Jensen,
1981).
Characters
Morphomelhc characters.— To develop morphometric charac-
ters for phylogenetic analysis, the following lengths were mea-
sured along the longitudinal axis of the fish; total length, preanal
length, head length, and eye diameter. Two vertical measure-
ments, head depth and body depth at anus, were meant to rep-
resent size and shape in the dorso-ventral direction. All mea-
surements were defined by Fuiman (1979). They were made
reasonably independent of one another by subtracting preanal
length from total length to yield peduncle length, and head length
from preanal length to yield tnank length. Peduncle length, trunk
length, head length, eye diameter, body depth, and head depth
comprised the basic morphometric characters.
(0
o
0
a
CO
0)
E
3
3-
2-
1-
\ZZ} Cypriniformes
I I Characiformes
^H Siluroidei
rXI Gymnotoidei
0.25
0.45
0.65
0.85
Yolk-Sac Shape (depth/length)
Fig. 68. Frequency distribution of yolk-sac shape for recently hatched
ostariophysan species.
Body dimensions of larvae are strongly influenced by allom-
etry (Fuiman. 1983b). Such measures cannot be expressed as
simple proportions, because the proportions are not constant
within a species throughout the larval period. The effect of size
on shape must be eliminated in comparisons of shape. Further,
any single measure which accounts for size in one taxon may
be an inappropriate measure of size in a distantly related taxon.
Within-group principal component analysis can be used to ex-
tract a size component, PCI (Humphries et al., 1981), that is a
linear combination of several variables, each containing infor-
mation on size and shape. Thus, PC 1 includes more information
on size than any single measure and is a better comparison across
taxa.
Univariate and multivariate methods of allometry relate dis-
tance measures log-linearly (Huxley, 1932; Jolicoer, 1963). Thus,
a within-species principal component analysis of the logarithms
of the six basic morphometric characters, based on the covari-
ance matrix, was performed to extract the size component (PCI ).
The extreme PCI scores for all taxa were compared and two
values (0.00 and 0.60), one near each end of the larval period,
were chosen as standard sizes for comparing morphometry. The
six morphometric measures were reconstructed for each of these
sizes by means of the regressions of the logarithm of the char-
acter on PC 1 . By selecting two sizes to compare, the phylogenetic
analysis included information on changing shape (allometry) as
well as static shape. The final 1 2 character values were recorded
as predicted lengths (in mm) for each morphometric measure
at each of 2 standard sizes. However, body depth at the anus
contained no discontinuous, phyletic variability. The final mor-
phometric characters were; (Characters 1 and 2) Peduncle length
(smaller and larger standard size, respectively), (3 and 4) Trunk
length, (5 and 6) Head length, (7 and 8) Eye length, (9 and 10)
Head depth. Three additional morphometric characters were
136
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
0.9
E
E
c
O)
c
(V
-J
0)
o
c
3
T3
(U
Q.
CI3 Cypriniformes
I 1 Characiformes
fTTTTI Siluroldel
— - Gymnotoidel
— I 1 1 1 1 1 1 —
0.0 0.2 0.4 0.6
Size (PCI score)
Fig. 69. Morphometric characters important in defining major groups of ostariophysan larvae. Shaded areas and individual Imes enclose all
regression-predicted values at two standard sizes (0.0 and 0.6) of a given taxon.
included: (11) Size at hatching (PCI score at total length for
hatching, based on the regression of PCI on the logarithm of
total length), (12) Size at complete finfold absorption (PCI score
at total length for complete finfold absorption, based on the
regression of PCI on the logarithm of total length), (13) Yolk-
sac shape (ratio of the greatest vertical length [depth] of the yolk
sac to its greatest horizontal length in recently hatched individ-
uals).
Meristic characters.— These include: ( 1 4) Preanal myomeres (all
myomeres at least partly anterior to a vertical line projected
from the anus, including an occipital segment) and (15) Postanal
myomeres (all myomeres entirely posterior to a vertical line
projected from the anus, including a urostylar segment).
Missing myomere data for Hoplias malabancus were taken
from vertebral counts made from radiographs of adults (UMMZ
66435). The one-to-one ontogenetic relationship of myosepta
to neural spines in monospondylous fishes (Lauder, 1980) per-
mitted estimation of myomere number from vertebral number
only by inclusion of myomeres for an occipital segment, a uro-
stylar segment, and the four (five in siluroids) obscured We-
berian vertebrae (Fuiman, 1982a).
Ontogenetic characters. —Size, rather than chronological age, is
most closely related to development (Gerking and Rausch, 1 979).
Thus, total length at the onset of selected developmental events
was recorded. To compare these sizes among species with dif-
fering initial lengths and ranges of lengths for the larval period.
the logarithm of the hatching length was subtracted from the
logarithm of the length at a given event. This difference was
divided by the difference of the logarithms of length at complete
finfold absorption and at hatching (the criteria used here to
delimit the larval period). The resultant character was the per-
centage of the larval period that occurred prior to the event, an
estimate of relative timing of the event. When characters were
present at hatching or did not develop until after complete fin-
fold absorption they were coded as 0.00 or 1.10, respectively.
The following events were recorded: (16) Anal fin rays (first
distinct ray), ( 1 7) All median fin rays (all median fin rays present,
finfolds may persist, fin margins may be incomplete), (18) Yolk
absorption (complete absorption of yolk), (19) Head straight
(head free from yolk sac and not deflected downward), (20) Eye
pigment (first uniform pigmentation of retina).
Presence/absence characters— Presence (coded as I ) or absence
(0) of the following structures at any time during the larval
period was recorded: (21) Jaw teeth (teeth on the premaxilla,
maxilla, or dentary), (22) Adipose fin, (23) Caudal spot (con-
gregation of melanophores at the base of the caudal fin forming
a distinct spot), (24) Lateral stripe (melanophores on the mid-
lateral myoseptum forming a continuous, longitudinal stripe).
Phylogenetic results
The Wagner tree (Fig. 67, Table 28) contains 101 steps for
the 46 two-state factors ("characters"). Members of each major
Table 29. Distribution Statistics of Preanal. Postanal, and Total Myomere Number for Ostariophysan Larvae. Values are based on
means for each species.
Preanal myomeres
Postanal myomeres
' Including Cyprinidae, Calostomidae, and Cobiloidea.
Total myomeres
Taxon
species
Mean
Extremes
Mean
Extremes
Mean
Extremes
Cypriniformes'
52
29.4
18.5-38.8
12.4
7.0-18.8
41.7
32.8-50.7
Characiformes
4
25.5
17.8-32.7
15.2
8.0-20.0
42.0
36.1-50.0
Siluroidei
6
19.7
15.3-26.3
25.9
16.5-38.7
45.4
33.0-65.0
Gymnotoidei
1
17.3
45.9
63.2
FUIMAN: OSTARIOPHYSI
137
4-1
3-
co
.2
'5
0
a
CO
"?; 2-\
E
3
r//J Cyprinidae
I I Cobitoidea
^B Siluroidei
^ ^.
i
1
r
0.0
0.4
0.8
1.2
Barbel Formation
(onset as percentage of larval period)
Fig. 70. Frequency dislnbution of relative timing of" barbel forma-
tion in ostanophysan species. Cyprinids are represented by 1 0 barbelled
species, not all of which are discussed in the text.
taxon (Cypriniformes. Characiformes, Siluroidei) are placed near
one another, but larval characters are insufficient to demonstrate
the monophyly of characiforms or siluroids. The largest number
of primitive characters is found in Hoplias (Characiformes), but
the cypriniform lineage differs from Hoplias by only three char-
acter state changes (node 3). As suggested by Fink and Fink
(1981), the gymnotoids are most closely related to siluroids
(node 26).
The cyprinifoim lineage (node 4) is united by two unreversed
synapomorphies: an elongate yolk sac (Figs. 62A and 68) and
the absence of jaw teeth. Cypriniforms and characiforms uniquely
share large eyes at the larger standard size (PCI = 0.6; Fig. 69).
This character reverses to a plesiomorphous condition for the
siluriform lineage. Synapomorphies of siluriforms include a long
peduncle at the larger size (Fig. 69) (a unique state for the group,
except for a single reversal in Bagarius), short head at the larger
size (highly homoplasious), and small eyes at the smaller size
(PCI = 0.0; Fig. 69) (unique except for a reversal in Ictalurus).
The gymnotoid, Eigenmannia (node 28), expresses six auta-
pomorphies, two unique and two occurring in only one other
place on the tree. The uniquely derived conditions are a short
trunk at the larger size (Fig. 69) and numerous postanal myo-
meres (Table 29).
Several morphometric characters make valuable contribu-
tions to the phylogenetic reconstruction. The axial measure-
ments (head, trunk, and peduncle lengths) exhibit a clear trend
for increasing head and peduncle lengths at the expense of trunk
length through the cypriniform - characiform ^ siluroid -
gymnotoid phyletic sequence. A portion of the variation in pe-
duncle size is attributable to migration of the anus anteriad in
this phyletic sequence, as evidenced by decreasing preanal and
increasing postanal myomere counts (Table 29). However, the
remaining peduncle variation and that of the head length are
the result of allometry.
In Fink and Fink's (1981) study, a single character involving
the evolution of a new structure, a pair of barbels, conflicted
with their adult-based cladogram. Ontogenetic evidence sup-
ports their contention that the presence or absence of barbels is
a poor indicator of relationship in ostariophysans. An ontoge-
netic character for timing of barbel development (constructed
in the same manner as described earlier for other ontogenetic
characters) displays two distinct modes (Fig. 70). Cyprinids de-
velop barbels during the latter third of the larval period, often
after finfold absorption (i.e., as juveniles). Siluroids and co-
bitoideans' do so during the first third of the larval period,
sometimes prior to hatching. Although the sample size of cob-
itoideans is small, it appears that they develop barbels somewhat
later than the siluroids. Thus, although barbels are present in
adults of all three groups, there is an important difference in
these structures between the groups: heterochrony. That het-
erochrony is a major cause of evolutionary change was amply
attested by Gould (1977).
Heterochrony in barbels may be an important consideration
for classification within siluroids. The number of pairs of barbels
(usually counted in the adult stage) is an important character
for recognizing siluroid families. At least one pangasiid, Silonia
silondia. has been described in which the larvae have three pairs
of barbels (nasal, maxillary, and mandibular) that gradually be-
come smaller until only one pair of minute maxillary barbels
are present on the surface of adults (Karamchandani and Mot-
wani, 1956).
The phylogenetic analysis presented here is based on devel-
opmental characters. It shows general congruence with the most
thoroughly researched adult-based cladogram (Fink and Fink,
1981); however, larval characters alone are not as informative
as adult characters. Larval characters support the new idea that
gymnotoids are more closely related to siluroids than to char-
aciforms. Characiforms appear to be primitive ostariophysans
by virtue of the basal location of the relatively primitive char-
aciform Hoplias. The apparent paraphyly of characiforms and
siluroids is due to the lack of shared characters for each of these
groups and would be altered by the reasonable addition of the
numerous adult autapomorphies discussed by Fink and Fink
(1981). Once monophyly is demonstrated by adding adult char-
acters, Hoplias would probably occupy a basal position (with
respect to the other three characiforms examined here) on a
characiform lineage. However, the position of this lineage with
respect to that of the cypriniforms may or may not agree with
Fink and Fink's (1981) adult-based cladogram.
School of Natural Resources, S. T. Dana Building,
University of Michigan, Ann Arbor, Michigan 48109.
' Cobitoideans included here and in Fig. 70 were: Cobitidae— Bo/;a
.vafir/i; (Changjiang, 1976); Cobilis taenia (Chyung. 1961; Koblitskaia,
1981; Kokhanova, 1957; Kryzhanovskii, 1949; Kryzhanovskii et al.,
1951; Menasse, 1970); Mtsgurnus anguillicaudalus {Chyung,. 1961; Ko-
bayasi and Moriyana, 1957; Okada, 1960; Okada and Seiishi, 1938;
Suzuki, 1955, 1968); Homalopteridae— A'emac/jei/jis dorsalis (Kry-
zhanovskii, 1949).
Gonorynchiformes: Development and Relationships
W. J. Richards
THE Gonorynchiformes is a small group of fishes which have
been allied with the clupeiforms or salmoniforms and most
recently have been placed as a lineage, within the ostariophysan
group, which includes also the Cypriniformes, Characiformes,
and Siluriformes (Fink and Fink, 1981). The group is comprised
of seven genera classified in about four or five families. The
most widely known species is Chanos chanos Forsskil placed
in the monotypic family Chanidae. The Gonorynchidae is a
marine family of one genus Gonorymchus and several species
found in tropical waters of all but the western Atlantic and
eastern Pacific. The remaining twelve or so species are African
freshwater forms in the genera Kneria. Parakneria. Grassei-
chthys and Phractolaemus, which may represent two or three
families. The eariy life history of Chanos is very well known
because of the extensive culturing; Gonorymchus is poorly known.
The early life histories of the freshwater species are unknown.
Pellegrin (1935) notes that young specimens of Cromeria nt-
lotica have a superficial resemblance to young Albula. It is ap-
parent that this resemblance is to the shape of juveniles and not
to a leptocephalus stage. Several subsequent papers have erro-
neously reported that Pellegrin said that Cromeria resembled
larval Albula.
Development
The early life history of Chanos chanos, the milkfish, has been
described by Delsman (1926d, 1929b). Since Chanos is an im-
portant aquaculture organism, several recent papers have de-
scribed various aspects of development, among them the de-
scription by Liao et al. (1979) is the most complete. Miller et
al. (1979) provides a good account for separating them from
common marine larvae. To summarize, the eggs and larvae
superficially resemble clupeids and engraulids but differ in sev-
eral trenchant characters. The eggs as described by Delsman
( 1 929b) are spherical, 1 .2 mm in diameter, lack oil droplets and
have a weakly segmented yolk which may be similar to the
granular yolks seen in ostariophysans. Yolk-sac larvae have me-
lanophores scattered over the body and fin folds and a myomere
formula of 34 -I- 10 (preanal and postanal). As development
Fig. 71. Lateral and ventral views from top to bottom: Chanos chanos. 1 1.7 mm SL from Kumano, Tanegashima collected August 19, 1978,
drawn by J. C. Javech; and Gonorymchus abrevialus. 12.8 mm SL from R/V Shoyo Maru station 25, 35°05'N, 144°24.3'E. collected on November
10, 1963; drawn by J. C. Javech.
138
RICHARDS: GONORYNCHIFORMES
139
progresses, the melanophores collect along the dorsal and ven-
tral midlines of the trunk. In larvae 10-15 mm SL (Fig. 71)
pigmentation is variable with melanophores on the dorsal mid-
line varying from one to many and melanophores on the lateral
line varying from none to many. The ventral midline has a
continuous streak of melanophores in sharp contrast to clupeids
and engraulids which have melanophores laterally on each side
of the gut thus presenting two parallel streaks in ventral view.
The anal fin of Chanos originates beneath the dorsal fin as in
engraulids. In Hawaiian waters meristics separate Chanos from
Gonorynchiis and other clupeids and engraulids. Chanos has
40-46 vertebrae [44-46 according to Miller et al. (1979) and
40-45 according to Senta and Kumagai (1977)]. Dorsal rays are
14-16, anal rays 8-11, pectoral rays 17 and pelvic rays 10-12
(Miller et al., 1979).
Much less is known about the early life history stages of Gon-
orynchus. Furukawa (1951) described the larvae of G. ahbrev-
latus and illustrated 18 and 23 mm specimens. He based his
identification on dorsal (1 1-1 2) and anal (7-8) fin rays, vertebral
counts (55) and the posterior position of the dorsal and anal
fins. Hattori (1964) illustrated and briefly described a series of
G. ahbreviatus from 8.6 to 90.5 mm. He noted that the positions
of the dorsal and anal fins do not shift during development.
Mito (1966) illustrates two larval G. ahbreviatus. I examined a
series of G. abbreviatus specimens and one is illustrated here
(Fig. 7 1 ). The larvae resemble clupeids with the wide separation
of the dorsal and anal fin. Pigment occurs dorsally and ventrally
on the caudal peduncle and extends posteriad into the bases of
the procurrenl caudal rays. Internal pigment occurs above the
hindgut and behind the brain. A few external melanophores
are present on the top of the head. Additional external mela-
nophores appear with growth. These include a series which de-
velops as lateral spots increasing in number with growth. In a
few specimens examined a 15.9 mm larva had one spot and
these increased in number to 18. At 23 mm SL pigment also
appeared on the opercle and ventral rim of the orbit. The pelvic
fin is discernible as a bud in small larvae but fin rays are not
defined until 18 mm SL. A swimbladder is not discernible on
any of the specimens as it is in clupeids and Chanos.
Relationships
The relationships of the Gonorynchiformes have been dis-
cussed most recently by Fink and Fink (1981). They conclude
that this order is the sister group of the Otophysi (the taxon
which includes fishes with the Weberian apparatus). Chanos
and Gonorynchiis larvae more closely and superficially resemble
clupeoid larvae than any other group. This matter should be
thoroughly investigated when early life history aspects of the
freshwater species become better known. It will be interesting
to see if those larvae resemble the marine species or freshwater
Otophysi.
National Marine Fisheries Service, Southeast Fisheries
Center, 75 Virginia Beach Drive, Miami, Florida 33149.
Salmoniforms: Introduction
W. L. Fink
ORIGINALLY a major portion of the Protacanthopterygii
of Greenwood, et al. (1966), the order Salmoniformes is
now the only portion left in that group, and the former term
has ceased to have a useful function. This erosion of the Pro-
tacanthoptergyii has resulted from the search for and taxonomic
recognition of natural groups of primitive euteleosts, a practice
that has and is continuing to have profound effects on fish clas-
sification at all levels. This part of the symposium, concentrating
on the "salmoniforms," places its participants in the middle of
a continually changing set of problems, some of which have
been longstanding. One of the questions we address here is
whether the Salmoniformes as conceived by Greenwood et al.
is itself useful any more, and if not, what are the relationships
of the formerly included groups. In the years since it was delin-
eated, the Salmoniformes has undergone attrition, most notably
at the hands of Rosen (1973). Of particular concern to us is
whether there is one large monophyletic unit which can be called
Salmoniformes, as maintained by Rosen (1974), or whether
there are several units, as suggested by Fink and Weitzman
(1982), thus reciuiring us to modify our conclusions and clas-
sifications. The basic questions are these; (1) What are the re-
lationships of the Esocoidei (sensu Rosen, 1974), both to one
another and to other primitive euteleosts? (2) What are the
relationships of the Ostariophysi, (sensu Rosen and Greenwood,
1970)? Do these fishes lie above or below the Esocoidei in the
phylogeny? (3) What is the pattern of relationships among the
traditionally recognized "salmoniform" taxa, exclusive of the
Esocoidei and Ostariophysi? Is this a natural division? (4) What
are the phylogenetic relationships of and within the Argenti-
noidei (sensu Greenwood and Rosen, 1971)? (5) What are the
phylogenetic relationships of and within the Osmeroidei? (6)
What are the phylogenetic relationships of and within the Sal-
monidae? (7) Where does Lepidogalaxias belong? (8) What are
the interrelationships within the stomiiform fishes? (9) What of
the Myclophoidei, as recognized by Greenwood, et al. (1966)?
This "group" has been most recently addressed by Rosen (1973)
in his discussion on the Eurypterygii and Neoteleostei. Parts
of these groups overlap into areas covered by this particular
part of the symposium, such as placement of giganturids, and
other parts into non-"salmoniform" portions such as that on
myctophi forms.
In many ways this symposium is a report on the state of the
science of fish classification, will summarize current ideas of
relationships and, especially, will point to where the greatest
need for further research lies.
Museum of Zoology, University of Michigan, Ann Arbor,
Michigan 48109.
Esocoidei: Development and Relationships
F. D. Martin
THE Esocoidei consist of two families, Esocidae and Um-
bridae, with one and three genera respectively (Nelson,
1976). Table 30 lists all currently accepted species and gives
their geographic ranges. All recent classifications consider the
esocoids as members of the Salmoniformes (Greenwood et al.,
1966; Gosline, 1971; Rosen, 1974; Nelson, 1976; and others).
All esocoid fishes live in freshwater and occur in temperate and
arctic waters of the Northern Hemisphere. All species are pred-
atory with Esox being primarily piscivorous. They are distin-
guished from other salmoniform fishes by the lack of the meso-
coracoid, lack of pyloric caeca, a single rudimentary arch over
PUl, and a single uroneural (Rosen, 1974). Table 31 gives de-
velopmental features that characterize esocoid fishes and con-
trasts them with Salmonidae and Osmeridae.
Development
Eggs are demersal and adhesive in most species (Breder and
Rosen, 1966) but Esox niger eggs become buoyant at later stages
of development and are not adhesive after water hardening (Jones
et al., 1978). Eggs are of moderate size (1.0 to 2.2 mm usually)
(Jones et al., 1978) and are either scattered as by Esox or are
in nests as with Umbra and Novumhra (Breder and Rosen, 1 966).
Table 30.
Genera and Species of Esocoid Fishes and Geograph-
ical Ranges.
Esocidae
Esox
E. lucius
E. reicherli
E. masquinongy
E. niger
E. americanus
Umbridae
Novumhra
N. hiibbsi
Umbra
v. krameri
V. linu
U. pygmaea
Datlia
D. pectoralis
D. asmirabilis
Holarctic (Grossman in Lee et al.. 1980).
Amur River region of Siberia (Berg, 1948).
Eastern North America, primarily Great
Lakes and Upper Mississippi drainage
(Grossman m Lee et al., 1980).
East Goast drainage of North America, also
lower Mississippi drainage (Grossman in
Leeet al., 1980).
Eastern half of North America (Grossman m
Leeetal., 1980).
Olympic Peninsula of Washington State
(Meldnm m Lee el al., 1980).
Middle and lower Danube System and lower
Dniester River (Berg, 1948).
Southern Ganada and Gentral United States
(Gilbert m Leeetal., 1980).
Southeastern New York to Northern Florida.
mostly on Goastal Plain (Gilbert in Lee et
al., 1980).
Arctic and sub-Arctic Alaska and eastern tip
of Sibena (Rohde in Lee et al., 1980).
Amguema River basin of Siberia (Gheresh-
nev and Balushkin, 1980).
Multiple oil droplets occur with a unique set of movements
producing alternating clustering and dispersion as ontogeny pro-
ceeds (Malloy and Martin, 1982).
Larvae of nearly all species are known, and developmental
series have been described and illustrated. Figs. 72 and 73 show
representative larvae oi Esox and Umbra. Those described hatch
relatively undeveloped, with head flexed over and attached to
the large yolk sac; the eyes are unpigmented. In all species the
notochord is stout and reaches nearly to the margin of the caudal
finfold. During flexion the notochord extends well beyond the
developing hypurals and may form a separate lobe to the de-
veloping caudal fin until the hypurals are complete. In Umbra
and Esox the pectoral fin is the first to begin differentiation (but
not form rays) with the pelvic fin the last to develop fin rays.
All median fins differentiate more or less simultaneously with
caudal starting ditTerentiation slightly ahead of the others.
Changes in body form are gradual with no noticable point of
metamorphosis. Before fin differentiation is complete the body
Fig. 72. Development of Esox niger from hatching to juvenile.
Lengths arc total lengths. (From Mansueti and Hardy, 1967.)
140
MARTIN: ESOCOIDEI
141
Common Cardinal
Hepatic Vitelline Vein
-'Sublntestinal Vitelline Vein
Common Cardinal, /Hepatic Vitelline Vein
Fig. 73. Early yolk-sac and late yolk-sac larvae of Umbra pygmaea.
(From Wang and Kemehan, 1979.)
Heart
5.4 mm TL
-'Sublntestinal Vitelline Vein
Fig. 74. Schematic representations of the vitelline venous systems
of Esox americanus (upper) and Umbra pygmaea (lower)— based in
part on figure from Wang and Kemehan, 1979.
Table 31. A Comparison of Egos and Larvae of Esocoid, Salmonoid and Osmeroid Fishes. Unless otherwise noted information on
Umbridae and Esocidae taken from Malloy and Martin (1982).
Egg
Demersal
Adhesive
Oil droplets
Size
In nests
Embryo and yolk-sac larva
Head deflexed, adherent to yolk-sac
Eye pigmented at hatching
Vitelline circulation
Common cardinals
Hepatic vitelline vein
Sublntestinal vitelline vein
Sublntestinal v. v. forming rete
Hepatic v. v. forming rete
Larva
Vertebrae (myomeres)
Adipose fin
Dorsal origin over or behind anus
Notochord forming a urostyle
extending length of hypural
complex past hypurals
Juvenile and adult
Pyloric caeca
Anterior constriction of vertebra
Pharyngobranchial 1
Epurals
Hypurals
Neural spine on preural 1
Neural spine on preural 2
-1-
-1-
'-I-
' +
+
multiple
M-2.2mm
-1- or '-±
multiple
M.9-'3.4mm
'± (mostly -)
'•■ '-multiple
or * '-single
'M. 5-7.0 mm
-single or
" '^multiple
-1 mm
'-I-
1 _
' +
1 +
-I-
+
-t-
-I-
'32-""'42
-I-
+
+
+
+
+
'43-67
-I-
-I-
3.4 +
9. I0*_
9.10 +
9+ 10_
9.10_
9.10+ _
"•46-75
3.18 +
4.18_
2_
2.15 +
?
?
?
9
9
'55-70
0
0
"13-222
"0-11
20*« +
20*** _ +
20 _
20_
-
-
+
-1-
0-2
2
"2 or 3
"2 or 3
5 or 6
6
"7
"6
-1-
+
"-or reduced
"-or reduced
fully
fully
reduced or
"fully
developed
developed
not
developed
• Present bul does not run on surface of yolk sac.
•" In Novumbra and Daltia only present in midabdominal region of juv
••• When present ther« is also posterior constriction.
' Breder and Rosen, 1966.
'Cooper. 1978
' Rajagopal. 1979.
^ Watling and Brown, 1955.
• Baugh. 1980
• Jones ctal.. 1978.
'Carbine. 1944.
" Uach. 1923.
'Soin. 1966.
'"Kunz. 1966.
' Bigelow and Schroeder, 1963.
' Fuiman. 1982b.
'Nelson. 1972.
' Auer, 1981-
^ Yanagawa. 1 978.
" Scott and Crossman. 1973.
' Hart. 1973.
■Nagiec, 1979.
' Greenwood and Rosen. 1 97 1 .
•Cavender, 1969.
142
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
form is basically that of the adult. Guts are simple with no
elaborations in all species. At hatching Umbra has a shorter gut
and fewer myomeres than Esox and this is reflected in there
being 5 myomeres between the yolk sac and the anus in newly
hatching U. pygmaea and 12 in E. americanus (Malloy and
Martin, 1982).
Relationships
Malloy and Martin (1982) point out three ontogenetic char-
acteristics shared by Esox and Umbra, which indicate close
relationship. The position of the heart at the time of formation
is on the yolk sac anterior to and left of the head. All other fish
for which position of the forming heart is noted have it forming
under the head in the pericardial cavity or, as in the Atherini-
formes, near the midline and anterior to the head. The yolk-sac
circulatory pattern consists of paired simple common cardinals,
a posterior rete formed by the subintestinal vitelline vein and
paired or single hepatic vitelline veins which enter the rete before
the subintestinal vitelline vein joins the common cardinals at
the heart (see Fig. 74). This differs from all other salmoniform
fish for which the pattern is described (Kunz, 1 964; Soin, 1 966).
The oil droplets go through a predictable series of clustering and
dispersion. Oil droplet movement of this sort has only been
documented previously by Ahlslrom ( 1 968) for bathylagid smelts
of the genera Bathylagus and Leuroglossus.
McDowall (1969) recognized a salmonoid-osmeroid-esocoid
lineage but states "Where esocoids fit into this series of sub-
orders and families is not clear to me." Rosen (1973) likewise
considers the esocoids and salmonoids to probably be closely
related but considers this alignment to be provisional. Fink and
Weitzman (1982), in contrast, state that they find no evidence
to consider the esocoids closely related to the other Protacan-
thopterygii (sensii Rosen, 1 974), which are the Agentinoidei and
Salmonoidei (including the Salmonoidea plus Osmeroidea).
Fink and Weitzman list the esocoids as sedis mutahilis at the
euteleostean level or as the sister group to all other euteleosts.
Soin (1980), on the basis of egg development patterns, feels that
the esocoid fish are incorrectly placed as a suborder of the Sal-
moniformes, however he gives no guidance as to correct place-
ment. While the ontogenetic evidence presented in Table 30 is
not conclusive it suggests that there is a large difference between
the esocoids and the Salmonoidei and this is consistent with the
opinions of Fink and Weitzman.
The vertebrae of Umbrids have a pronounced anterior con-
striction, giving them an asymmetrical appearance, however
Novumbra and Dallia show this characteristic only while young
and most noticeably in the mid-abdominal region. In Esox the
vertebrae are either unconstricted or are constricted both an-
teriorly and posteriorly so that they appear symmetrical (Cav-
ender, 1969). Other differences between the Esocidae and the
Umbridae are seen in the Umbridae having nine or fewer bran-
chiostegals, fewer infraorbitals, no supratemporals or intercalars
and usually fewer than 41 vertebrae (Wilson and Veilleux, 1982).
Chesapeake Biological Laboratory, University of
Maryland, Box 38, Solomons, Maryland 20688.
Salmonidae: Development and Relationships
A. W. Kendall, Jr. and R. J. Behnke
SALMONIDS (whitefishes, ciscoes, grayling, trout, and salm-
on) are highly important in terms of aesthetic appreciation,
commercial and recreational value, and scientific study. Studies
of the development of salmonids from hatching until the time
of yolk depletion, and of the relationships among subfamilies
and genera have been largely neglected [see review of systematics
by Dorofeyeva et al. (1980)] despite the large body of literature
on early embryological development and relationships among
species and populations. Salmonids all spawn in fresh or brack-
ish water, some are anadromous while others are strictly fresh-
water. The family is composed of about 10 genera in three
subfamilies: Coregoninae, Thymallinae, and Salmoninae (Table
32) (Nelson, 1976).
Along with a precise homing ability, salmonids tend to form
genetically isolated populations. They seem to be able to occupy
new niches and habitats as these become available in the cold
temperate parts of the Northern Hemisphere. One result of this
adaptability is the existence of taxonomic problems mainly at
the species-population levels (Utter, 1981).
Development
Post-hatching development of salmonids has been little stud-
ied (Table 33), and only a superficial analysis of comparative
developmental stages has been attempted (Soin, 1980). Thy-
mallus and the salmonines share apparently advanced features
of development such as large yolk sac with an extensive vitelline
circulatory system and development of rather uniform intense
pigment, while coregonines develop larvae that are more typical
of other freshwater fishes (Faber, 1970). Thymallus seems inter-
mediate between the coregonines with a "normal" larval stage
and the salmonines in which the larval stage is largely bypassed
(the young have fully formed fins by the time the yolk is ab-
sorbed). Parr marks (vertical blotches or bars of pigment over
the trunk of juveniles) are present in all salmonids except Cor-
egomts a.nd StenodushuX are not seen injuveniles of other fishes.
Norden (1961) incorrectly considered the early stages of Core-
goniis artedii as figured by Fish (1932) to be similar to those of
Thymallus arclicus. He also stated that "the development of
KENDALL AND BEHNKE: SALMONIDAE
143
Table 32. Characters that vary among the Salmonid Subfamilies.
Subfamily
Character
Coregoninae
Thymallinae
Satmoninae
General
Genera
Coregonus, Prosopium. Steno-
Thymallus
Brachymystax. Hucho, Salvelinus.
dus
Salmo, Parasatmo, Oncorhvnchi
Species
30
4
32
Habitat
freshwater, few anadromous
freshwater
freshwater and anadromous
Egg size
1.8-3.7 mm
2.5 mm
3.7-6.8 mm
Diploid chromosome num-
64-82
102
52-92
bers
Dorsal fin rays
10-15
17-25
8-12
Dentition'
Tooth character
narrow, sharp, 2-3 sections
uniform in size
vary in size
Maxillary
toothless
toothed
toothed
Dentary
minute teeth restricted to ante-
narrow, teeth of uniform size
numerous teeth of varying size all
rior end
all along bone
along bone
Vomer
small and toothless (except in
Stenodus and some Corego-
nus)
small, with teeth
long, with teeth
Premaxillary
small
large
large
Caudal skeleton-
Epurals
3
3
2-3^
Stegural
little developed
little developed
well developed
Neural and hemal spine ex-
little
moderate
large
pansion
Urodermal
present
absent
absent
Neural spine on PU,
absent
absent
present
Neural spine on PU,
not fully developed
not fully developed
fully developed''
Cranial osteology'
Orbitosphenoid
present
absent
present
Suprapreopercular
absent
absent
present
Panetals meet at midline
yes
yes
no
Hypethmoid
present
absent
usually absent
Basisphenoid
usually absent
present
present
Uppermost orbital'
present
present
absent
' Vladykov (li)70).
■'Cavender (1970)-
' Nordcn (1961 1,
* Some vanation withm Salmonmae in these Iwo characters. Those with 2 cpurais usually have most extensive neural spine development.
^ Sometimes erroneously termed dcrmosphenotic; sometimes present in Salmoninae; see Behnke (1968, p 9-10).
the young grayling has much in common with that of both the
coregonines and salmonines" (Norden, 1961:743).
Among the coregonines, larvae of Prosopium (Faber, 1970;
Auer, 1982), L«/c;c7!r/!V5 (Fish, 1932; Faber, 1970; Auer, 1982),
and Coregonus (¥'\%\\. 1932; Faber, 1970; Auer, 1982) have been
illustrated and briefly described. All show similar larval mor-
phology (Fig. 75). They are rather slender with a long preanal
finfold— the yolk being confined to the anterior trunk region.
The yolk-sac length is <35% total length (TL), eye diameter is
<7% TL, and body depth at anus is usually < 10% TL (Auer,
1982). The yolk is exhausted before any of the fins, except the
caudal, possess full complements of rays. Prosopium eggs have
multiple oil globules, while Leucichthys and Coregonus eggs
have a single oil globule (Auer, 1982). Pigment in preflexion
and flexion larvae is mainly associated with the dorsal and ven-
tral midlines. Later, the body becomes more uniformly pig-
mented. Prosopium develops parr marks during the juvenile
period. Larvae oi Stenodus are undescribed and they may differ
from those described above, since adults of this genus appear
quite divergent from the others in this subfamily.
Early development of Thymallus thymallus has been fully
described (Penaz, 1975). They hatch with a large, anteriorly
placed yolk sac that is covered by a rather extensive vitelline
circulatory system, and the preanal and postanal finfolds are
about equal in length (Fig. 75). The yolk sac is exhausted during
notochord flexion and by that time some fin rays have developed
in all of the fins. The larvae are rather heavily pigmented during
this period. When the fins have developed their adult comple-
ment of rays, the fish appear like juveniles and parr marks begin
to form.
Early development of all the salmonine genera and most sub-
genera is known, although several are inadequately described
(Table 33). Described development of all salmonines is quite
similar (Figs. 76, 77). Their eggs are among the largest of all
teleosts. They all hatch with large yolk sacs and well developed
vitelline circulatory systems. The preanal finfold is shorter than
the postanal finfold (except in Hucho where they are about
equal). The preanal finfold extends somewhat down the poste-
rior of the yolk sac in Oncorhynchus. The notochord is slightly
flexed and some caudal rays are present. Yolk-sac length is
144
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 33. Meristic Values and References to Descriptions of Larvae of Salmonids. Total reported ranges of meristic values are given,
although the extremes of the ranges may be rarely observed.
Subfamily
Genus
Subgenus
Ranges of meristic values
References with
illustrations of flexion
stage larvae
Verte- Dorsal
brae' fin-
Pec-
Tolal
Lateral
Branch i
toral
Pelvic
gill
line
ostcgal
fin
fin
rakers
scales
rays
Primar\'
source
Coregoninae
Stenodus
Prosopium
Coregonus
Leucichlhys
Coregonus
Thymallinae
Thymallus
Salmoninae
Brachymystax
Hucho
Hucho
Parahucho
Salvelinus
Sahelinus
Baione
Crislivomer
Salmo
Salino
Salmothymus
Acantholingua
Plalysalmo
Parasalmo
Oncorhvnchus
64-69 12-19 15-18 16-17 11 19-24 90-110
Faber (1970), Auer( 1982) 50-65 10-15 10-14 13-18 9-12 11-44
9-12 Scott and Cross-
man (1973)
50-108 6-10 Scott and Cross-
man (1973)
Fish (1932); Faber (1970). 50-67 8-15 9-16 13-18 8-13 21-64 58-110 7-10 Scott and Cross-
Auer(1982) man (1973)
Fish (1932), Faber (1970), 55-64 10-13 9-14 14-17 11-12 15-78 70-102 6-10 Scott and Cross-
Auer(1982)
Penaz(1975)
Smol'yanov (1961)
Balon(1956)
Balon (1980)
58-62 17-25 11-15 14-16 10-11 16-33 81-103 7-9
man (1973)
Scott and Cross-
man (1973)
58-62 12-15 11-14 15-18
64-71 12-14 11-13 15-18
57-62 12-14 12-14 14-17
9-10 20-30 120-150 10-13 Behnke(1968)
and original
10 10-17 120-150 9-12 Behnke(1968)
and original
9 14-20 110-120 9-12 Behnke(1968)
and original
57-71 10-12 8-10 14-16 9-11 11-51 105-152 10-15 Scott and Cross-
man (1973)
Balon (1980), Auer( 1982). 57-62 10-14 9-13 11-14 8-9 14-22 110-130 9-13 Scott and Cross-
Martinez (1983) man (1973)
Fish (1932), Balon (1980). 61-69 8-10 8-10 12-17 9-10 16-26 116-138 10-14 Scott and Cross-
Auer(1982)
Auer(1982). Martinez
(1983)
Auer(1982), Martinez
(1983)
Auer(1982)
man (1973)
54-62 10-15 8-13 12-16 9-10 14-25 100-130 10-12 Behnke(1968)
and original
56-60 13-15 11-13 12-14 9-10 25-32 100-115 10-12 Behnke(1968)
and original
52-59 11-13 10-12 11-13 9-10 18-22 95-110 9-11 Behnke(1968)
and original
57-59 13 11 14 9 23-24 109-110 10-11 Behnke(1968)
and original
55-67 8-12 8-12 11-17 9-10 14-28 100-150 9-13 Scott and Cross-
man (1973)
61-75 9-16 12-19 11-21 9-11 18-43 120-160 11-19 Scott and Cross-
man (1973)
Overall ranges
50-75 8-25
-19 11-21 8-13 10-78 50-160 6-19
' Vanalions exist in the literature in how many of last 3 upturned vertebrae are counted; some authors omit the last 3 upturned vertebrae.
' Includes rudiments where specified. A variation of 2-3 rays may result from different methods of counting (whether unbranched or rudimentary rays are included).
>35% TL, eye diameter >7% TL, and body depth at anus
usually > 10% TL (Auer, 1982). Pigmentation is unifoimly heavy
at hatching or later in the yolk-sac stage. The median fins de-
velop rays before the paired fins. By the time the yolk is absorbed
the finrays have completed foimation and the fish takes on a
juvenile appearance. Thus, the yolk remains a source of nutri-
tion throughout the larval stage.
Relationships
Although salmonids are considered to be living representa-
tives of the basal stock from which euteleostean evolution pro-
ceeded, there is no clear consensus on their relationships to other
fishes. Since there are differing opinions on the relationships
between the major teleostean lineages (i.e., the divisions of
Greenwood et al.; 1966), it is difficult to select representatives
of outgroups to compare with the salmonids. Recent studies
(Rosen, 1974; Fink and Weitzman, 1982; Fink, this volume)
have pointed out that the Protacanthopterygii and even the
Salmoniformes are probably not monophyletic taxa. The sal-
monids along with the galaxioids, osmeroids, and argentinoids,
may form a group (Salmonae) that is the primitive sister group
of the neoteleostei. However, the relationships among these
groups is not clear, and the salmonids may be closer to the
neoteleostei than to these other groups with which they have
frequently been aligned (Fink and Weitzman, 1982; Lauder and
Liem, 1983; Fink, this volume). Some primitive teleost traits
KENDALL AND BEHNKE: SALMONIDAE
145
Fig. 75. Flexion stage larvae of: (A) Coregonus (Leucichthys) artedii (17.5 mm); (B) Coregonus (Coregonus) clupeaformis (18.5 mm); (C)
Thymatlus ihymallus {\6.0 mm). A and B from Fish (1932), C from Penaz (1975).
Table 34. Characters that vary among the Coregonine Genera and Subgenera (sg) mainly from Norden (1961) and Cavender
(1970).
Coregonu.
Prosopium
Character
Coregonus (sg)
Leucichthys (sg)
Stenodus
Species
8
17
1
1
Habitat
Some occasionally anad-
Several anadromous
Freshwater
Anadromous
romous
Basibranchial plate
Absent
Absent
Present
.Absent
Parietal bones meet along
Yes
Yes
Yes
No: narrowly separated
midline
Postorbitals in contact with
Yes
Yes
Yes
No
preopercle
Parr marks
Absent
Absent
Present in some
Absent
Flaps between nostnls
2
2
1
2
Mouth size
Small
Moderately large
Small
Large
Teeth
Weak or none
Weak or none
Weak or none
Many, small
Mouth position
Subterminal
Supenor or terminal
Subterminal
Terminal
Vomer
Small, toothed in
some
Small, toothed in
some
Small, toothless
Large, toothed
First supraorbital
Moderate
Moderate
Short
Long
Supraethmoid
Short
Short
Long
Short
146
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 76. Flexion stage larvae of: (A) Brachymystax lenox(\1.2 mm); (B) Hucho (Hucho) hucho (20.8 mm); (C) Salvelinus (Sahelinus) alpinus
(19.8 mm); (D) Sahelinus (Cnslivomer) namaycush (approx. 20.4 mm). A from Smoryanov (1961), B from Balon (1956), C and D from Balon
(1980).
Fig. 77. Flexion stage larvae of; (A) Sahelinus (Batone) fontmalis ( 1 4.0 mm); (B) Parasahno gairdnert ( 1 4.0 mm); (C) Parasalmo darki (14.2
mm); (D) Salmo trutta (14.0 mm); (E) Oncorhynchus tshawytscha (25.0 mm). A-D from Martinez (1983), E original.
KENDALL AND BEHNKE: SALMONIDAE
147
148
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
large mouth w 7i3
many small teeth n 753
^large toothed vomer
^ N 720
no teeth on vomer
N 720
/
mostly:
small mouth n 7S3
small teeth on vomer
of young only n=720
parr marks absent n 743
two flaps between nostril n 71
enlarged first supraorbital n 71
Joss of basibranchial plate n 71
/ \\
N 727 slightly notched ethmoid
cartilage
light spots 1
Palatine vomer strong ascending pre-
teeth form z^ maxillary process J n
j" shaped "C gap between palatine I
band n 753^v vomer teeth 1
Thymallinae
N 72S notched ethmoid
\ cartilage
N 732 palatine vomer
teeth form a "T"
/
Coregoninae
N 679 no teeth on maxillary
c 9 one urodermal
c 11 small neural spine on PU^
N-753 < 16 dorsal rays
N-750 general loss of teeth
c 11 small neural spine on PU-,
N-679 > 16 dorsal rays
N-e79 no orbitosphenoid
increase in size and amount of yolk in egg
bypass larval stage
Salmonidae
100)
• N 739
• •N 752
♦IM 743
*N 752
• N 752
• *C 27
• • N 752
• •rj 752
• *N 752
• "N 752
• •C27
«»C 11
• *c n
*?C9
Tetraploid karyotype ( 2n chromosomes - '
Axillary pelvic process
Three upturned caudal vertebrae ( two ural centra)
Parr marks in juveniles
Three post cleithra
Mesopterygoid toothless
Last four hemal spines and parhypural fit together ( peg and socket)
Adipose fin present
Oviducts incomplete or absent
Mesocoracoid present
Opisthotic present
Principal caudal rays = 19
Three epurals
Full neural spine on PU,
Two hypurais (ventral) on U^, 4 hypurals (dorsal) on U, .
1 long, 2 short uroneurals
N 726 blunt pointed ethmoid cartilage
N 732 gap between palatine vomer teeth
M '53 no ascending premaxillary process
N 753 postorbitals contact preopercular
N 753 opisthotic touches prootic
N 728 reduced dorsal fontanelles in adult
well developed stegural
expanded caudal neural and
hemal spines
neural spine on PU.
large neural spine on PU,
N-679 parietals separated by
supraoccipital
N 739 small scales ("> 100 in
lateral Ime)
N-679 suprapreopercular present
N 731 curved preopercular
N 736 dorsal rays < 16
e-9 reduction or loss
of hypethmoid
B - Behnke 1968
N - Norden 1961
C - Cavender. 1970
H = Hol£ik,1982
(number refers to page
in above references)
* Salmonidae (synapomorph for family)
"" Salmonoidei (synapomorph for suborder)
■•• Shared primitive (plesiomorph) character
with other "primitive" teleosts
Fig. 78. Hypothesis of relationships among extant saimonid genera. Groupings and branching points are based largely on a consensus of recent
literature and are not the result of a strict cladistic analysis.
possessed by salmonids include lack of oviducts, presence of
abdominal pores, and three upturned caudal vertebrae sup-
porting the hypurals. Salmonids are autapomorphic with about
twice the DNA content of other '^salmoniform" families, ap-
parently the result of having a common tetraploid ancestor. The
salmonids possess an adipose fin, a mesocoracoid, pyloric caeca,
and the vestige of a spiral valve intestine. The gill membranes
extend far forward free from the isthmus and there is a pelvic
axillary process. Two shared derived features of the salmonids
and neoteleostei are: 1) the articulation of both the basioccipital
and exoccipital with the first vertebra, and 2) the presence of a
medial cartilage between the ethmoid and premaxilla (Fink and
Weilzman, 1982).
Although it is not possible at present to perform a meaningful
cladistic analysis of the salmonids, some evidence is available
in the literature which can contribute to such an analysis (Fig.
78). Cavender (1970) compared the osteology of leptolepids.
extinct fish thought to represent the basal teleost condition, with
that of the salmonids. He found several characters that indicated
1) that the salmonids are monophyletic, and 2) how the three
subfamilies of salmonids are interrelated. The coregonines ap-
peared to be most similar to the leptolepids, the thymallines
more derived than the coregonines, and the salmonines more
derived than the thymallines. Reshelnikov (1975). on the basis
of several types of characters, suggested elevating the subfamilies
to familial status.
Coregoninae contains about 30 species in three genera. They
are mainly freshwater, and produce rather small eggs, compared
to those of the other two subfamilies. They share several ad-
vanced characters with the other subfamilies, indicating that
salmonids are monophyletic, but lack a number of advanced
character states possessed by the other two subfamilies, as these
branched oflTafter the coregonines. Within the coregonines, Pro-
sopium seems least diverged (Table 34). Sienodus shows several,
possibly secondarily derived character states concordant with
feeding on large active prey (expanded dentition, large mouth).
Coregoniis, which seems to be a sister group to Stenodus. is
separated into two subgenera: Leucichthys with adaptations for
plankton feeding, and C orego nus y^h\ch are mainly benthic feed-
ers.
Thymallinae contains one genus, Thymallus, with about four
species in freshwater of the colder parts of the Northern Hemi-
KENDALL AND BEHNKE: SALMONIDAE
149
Table 35. Characters that vary among the Salmonine Genera.
Characters
Brachvmyslax
Hucho
Salvetinus
Salmo'
Parasalmo^
Oncorhynchus^
Subgenera
Hucho. Para-
hucho
Sahelinus,
Baione. Crisli-
vomer
Salmo, Salmo-
ihy/mts. Acan-
iholingua.
Ptatvsahno
Species
2
3-5
8
8
5
6
Habitat
freshwater
freshwater and
freshwater and
freshwater and
freshwater and
usually anadro-
anadromous
anadromous
anadromous
anadromous
mous
Mouth size
small
large
large
large
large
large
Teeth on shaft of
no
no
no
yes
yes
yes
vomer
Palatine-vomer-
U-shaped band
U-shaped band
teeth narrowly
teeth narrowly
teeth narrowly
teeth widely sepa-
ine teeth
separated
separated
separated
rated
Postorbitals con-
no
no
no
no
no
yes
tact preopercle
Supraethmoid
long, with nu-
broad, with nu-
long, with nu-
notched poste-
notched poste-
deeply notched
shape
merous poste-
rior projec-
tions
merous short
posterior pro-
jections
merous poste-
rior projec-
tions
riorly
riorly
posteriorly
Ascending pre-
intermediate
intermediate
extended and
intermediate
intermediate
none
maxillary pro-
sized
sized
well developed
sized
sized
cess
Opisthotic touch-
no
no
no
no
no
yes
es prootic
Dorsal fonta-
persistent
persistent
persistent
persistent
persistent
reduced in adult-
nelles
Egg size
4-5 mm
large
4-5 mm
5-7 mm
large
large
Diploid chromo-
92
84
78-84
56, 80-82'
56-70
52-74
somes
Dark spots-light
yes
yes
no
yes"
yes
yes
background
' There is lack of agreement on the relationships between these laxa; e.g., some consider Parasalmo a subgenus in Salmo. while others would also consider Oncorhynchus a subgenus of Salmo.
* Retained in O. ma.sou
* Salmo salar has 56-60 diploid chromosomes-
* Salmo marmoratus and S platycephalus have no dark spots.
sphere. They have several character states that seem advanced
over those seen in coregonines. They are moderate-sized, gen-
eralized insectivores (Table 32).
Salmoninae contains four to six genera, depending on opin-
ions over the relationships among the species in Salmo, Par-
asalmo, and Oncorhynchus (Table 35). These seem to be the
most advanced of the salmonids, and share several character
states that are derived compared to the other two subfamilies
(Table 35). Holcik (1982) presented evidence which suggests
that the genera Hucho, Brachymysla.x, and Salvelinus form one
lineage; Parasalmo and Salmo another; and Oncorhynchus a
third. Salmonines are mainly active predators and most tend
toward an anadromous life histoi7.
Early life history and developmental information should con-
tribute to the rigorous analysis of characters that will be required
to validate the foregoing hypotheses about relationships. Such
information is not presently available in the literature, but should
be readily obtainable, since so many of these fishes are routinely
reared in laboratories and hatcheines. Developmental infor-
mation seems particularly promising in this family, since a wide
range of the life history patterns are present and larvae can be
superficially grouped according to their representative subfam-
ilies.
(A.W.K.) Northwest and Alaska Fisheries Center, 2725
MoNTLAKE Blvd. E., Seattle, Washington 98112 and
(R.J.B.) Department of Fishery and Wildlife Biology,
Colorado State University, Fort Collins, Colorado
80523.
Southern Hemisphere Freshwater Salmoniforms: Development
and Relationships
R. M. McDowALL
SEVERAL family-level groups of diadromous salmoniform
fishes are found in cool-temperate southern hemisphere
fresh waters, forming an obvious ecological counterpart to the
northern cool-temperate Salmonidae, Osmeridae, Plecoglossi-
dae, Salangidae, etc. With the exception of a single species, in
a high elevation lake in New Caledonia, they are all south of
about latitude 28°S. They occupy all of the main land masses
(Australia, New Zealand, South America, South Africa) and
some of the more distant southern islands (Lord Howe. Chat-
hams, Aucklands, Campbell, Falklands). Diagnostic familial and
generic characters are listed in Table 36.
Familial arrangement of these fish varies from including all
in a single purportedly monophyletic family Galaxiidae (Nelson,
1972), through two families in separate sub-orders (Rosen, 1974)
to four families in one or two suborders. There are two obvious
and widely accepted familial groupings: Galaxiidae— Aplochi-
tonidae and Retropinnidae — Prototroctidae (McDowall, 1969).
The most recent view (Fink and Weitzman, 1982) suggests that
these four family level taxa are possibly all of osmeroid deri-
vation agreeing with my own evaluation (McDowall, 1 969), and
in contrast with Rosen (1974 — he links galaxiids and aplochi-
tonids with salmonoids; retropinnids and prototroctids with
osmeroids). The southern taxa are all clearly primitive prota-
canthopterygians of salmoniform type. Beyond that little can
be said other than that a further search of additional character
complexes is needed to clarify relationships.
Within-family relationships are little studied. Three of the
southern families (Retropinnidae, Prototroctidae, Aplochiton-
idae) can be dealt with more simply than the fourth (Galaxiidae).
Retropinnidae (Australia and New Zealand— see McDowall,
1979). — four species in two genera: Present state of knowledge
does not permit explicit recognition of affinities. Elongation of
the alveolar process in the premaxilla of Stokellia anisodon is
an advanced character which leaves three species of Retropinna
with the primitive condition (alveolar process short, maxilla
sometimes toothed). Stokellia also has unossified gill rakers (an
"advanced" but "loss" condition) and high scale count (100
compared with 70 or less in Retropinna— which is the derived
condition?)
Prototroctidae (Australia and New Zealand— see McDowall.
1976).— Two species in one genus. Two congeneric species pose
no phylogenetic problems. The only significant question to ask
is "How do these species relate to the Retropinnidae?" Answers
to this question have not yet been sought.
Aplochitonidae (Tasmania and South America— see McDowall.
197 la).— Three (perhaps four) species in two genera. Mono-
phyly of the Aplochitonidae (Aplochiton and Lovettia) should
not be assumed. Inclusion of Lovettia in the Gala.xias-Aplo-
chiton assemblage is supported by characters in Table 36 but
Lovettia has such reduced osteology that a search for characters
in other structural systems is needed before its relationships can
be clarified. Inclusion oi Lovettia in the Aplochitonidae is based,
in part, on history (it has always been there!) and in part, on
the fact that it is a "galaxioid" with the dorsal fin over the pelvics
and an adipose fin present (like Aplochiton and unlike Gala.xias).
Galaxiidae (.Australia. New Caledonia, Lord Howe. New Zea-
land. South America. South Africa). S'w genera with 37 species
distributed as follows: Gala.xias, 24— all areas but New Cale-
donia; Paragala.xias, 4— Tasmania; Neochanna, 3 — New Zea-
land; Gala.xiella, 3 — Australia; Brachygala.xias, 2 — South
America; NesogalcLxias. 1 —New Caledonia. This larger and more
complex family offers scope for phylogenetic analysis that has
had little attention.
Relationships
Previous studies of within-family relationships have been
based on morphological similarity (McDowall, 1 970), phenetics
based on muscle myogens (Mitchell and Scott, 1979), or den-
drograms derived from cluster analysis of morphometric or me-
ristic data (Campos, 1979). Johnson et al. (1981, 1983) have
sought to establish relationship on the basis of karyotypes and
multivariate analysis of morphometric and meristic characters
in the diverse Tasmanian fauna.
The only attempt at a "strictly phylogenetic" interpretation
of within-family relationships, by Rosen (1978), was based on
misinterpretation of character states and a limited perception
of variation in the family, and achieved nothing (McDowall,
1980). A broad and strictly phylogenetic analysis of galaxiid
inter-relationships is not yet available and probably depends on
examination of additional character complexes.
On the basis of out-group comparisons (all salmonoid— os-
meroid—galaxioid families have members that are diadromous)
it is my view that diadromy in the Galaxiidae is a primitive
character. It is represented in at least six species.
Diadromous species tend to be large and generalised in char-
acter, but with specific adaptations to habitats occupied during
freshwater life. Vertebral numbers are high (> 60) and ray counts
in pelvic (7) and caudal (16) fins very stable.
There are indications of close relationships with diadromous
stocks, e.g., Gala.xias maculatus seems likely to be a neotenous
derivative of some other diadromous galaxiid; distinctive ju-
venile colour patterns may relate G. argenteus to G. fasciatus
and perhaps G. truttaceus.
There are numerous landlocked populations of diadromous
species, and present interpretations are that several species are
derived by isolation following landlocking, e.g., G. auratus
(landlocked) derived from G. truttaceus (diadromous) in Tas-
mania; G. gracilis from G. maculatus in New Zealand.
Wholly freshwater species tend to be the more specialised
members, in which there is often dwarfing, reduced vertebral
counts, greater meristic instability, as well as the loss of the
distinctive marine juvenile stage. Some freshwater groups have
150
McDOWALL: SALMONIFORMS
151
not yet recognised origins within the diadromous stocks and
there is identifiable speciation related to known geo-tectonic
events. The relationships of some of the more distinctive species
groups— Neochanna (New Zealand), Galaxiella (Australia), and
including geographical outliers like Gala.xias zebratus (South
Africa) and Nesogalaxias neocaledonicus (New Caledonia)— re-
main obscure. Previous inclusion of Australian and South
American species in Brachygalaxias is ill-founded, on present
data, and confuses the understanding of relationships.
An interesting phylogenetic problem in the Galaxiidae in-
volves the diminutive Tasmanian Paragalaxias. with four species
in high elevation lakes that probably pre-date Pleistocene gla-
ciations. Paragalaxias is distinctive in having the dorsal fin
origin only a little behind the pelvic bases. In this regard it
resembles aplochitonids differing from all other galaxiids in which
the dorsal origin is close to the level of the vent/anal origin.
Thus is Paragalaxias a galaxiid in which the dorsal fin has
migrated forwards, the resemblance to Aplochiton being con-
vergent or is it an aplochitonid in which the anterior dorsal fin
position is primitive but in which the adipose fin has been lost?
Examination of additional character complexes in which gal-
axiids and aplochitonids differ is needed to clarify this question.
The preceding discussion makes it evident that relationships
between and within the southern diadromous salmoniforms re-
main in need of clarification. Only the Galaxiidae is large and
diverse enough to provide fertile ground for a study of within-
family phylogeny. In all the families, species and characters are
conservative in nature and lack distinctive or extreme speci-
alisation. Inter-specific differences tend to be expressed as changes
in meristic characters (like vertebral and fin ray counts), often
to presence/absence character states (pyloric caeca, canine teeth)
and sometimes to distinctive and stable differences in colour
patterning. There are few readily evident characters that are
indicative of major phyletic lineages. Possibly investigation of
laterosensory papillary rows will be informative. At present,
establishment of phylogenies appears difficult. A study of re-
lationships using DNA hybridisation techniques (Sibley and
Ahlquist, 1981) is at present in early planning stages.
Life History Patterns and Reproduction
In general life history patterns are understood although details
are sparse. There are broad similarities in patterns.
Retropinnidae.— Aspects of early life history have been de-
scribed by Milward {1966 — Retropinna sewon/— Australia),
Jolly (1967-«. retropinna— N.Z.) and McMillan (1961— Sto-
kellia anisodon—N.Z.). The eggs are tiny— 0.5 to 0.6 mm in
lacustrine R. retropinna, 0.95 mm in R. semom. They are de-
mersal and adhesive, spherical, without distinctive features. They
are a pale straw colour. They are deposited on sandy bottoms
in lower river reaches or estuaries (around lake shores in land-
locked populations), where development occurs; development
is relatively slow ( 10-20 days) and description of development
shows nothing distinctive (Fig. 79). Newly hatched larvae in
some species go to sea. In others they are lacustrine or riverine.
Larvae at hatching are small (2-5 mm), very slender and elon-
gated, the yolk sac with a single oil globule, and situated ante-
riorly beneath the opercular openings/pectoral fins. The gut is
long, the vent at about 70% of length. A continuous finfold
encompasses the trunk. Pectoral fin buds are present. Newly
hatched larvae are positively phototropic. Pigmentation and
later development are undescribed. Juveniles from a summer-
Table 36. Character States in Principal Genera of Southern
Freshwater Salmoniforms. (* except Paragalaxias; + present, - ab-
sent; u uniserial; m muUiserial; 1 parhypural + hypurals; 2 tubercles in
Lmetlia may not be comparable with others). Figures are "usual" al-
though variants are known. The divergent galaxiid genera are excluded
(Paragalaxias, Galaxiella, Neochanna, etc.).
Proto-
Retro-
Galaxi-
Irocli-
pinni-
idae
Aplochilonidae
dae
dac
Ga-
Lovel-
Prolo-
Retro-
Characters
laxtas
Aplochiton
tia
Iroctes
pinna
Dorsal fin
Over pelvics
x
X
X
X
Over anal
X*
X
Adipose
-
+
+
+
+
Scales
-
-
-
+
+
Homy keel
-
-
-
+
+
Cucumber odour
-
-
-
+
+
Pyloric caeca
4-
-1-
+
-
-
Vomerine shaft
long
-
long
short
short
Vomerine teeth
-
-
-
+
+
Basi branchial
-
-
-
-1-
+
teeth
Palatine teeth
-
-
-
+
+
Mesopterygoidal
u
u
u
m
m
teeth
Extrascapular
-
-
-
+
-1-
Ectopterygoid
~
slender
splint
~
+
+
Coracoid-cleithrum
-
-
-
+
+
process
Posterior pubic
-
-
-
+
-1-
symphysis
Pubic foramen
-
-
-
+
+
Caudal skeleton
1 +5
1 + 5
1 +5
1 +6
1 +6
Branched caudal
14
14
14
16
16
rays
Nuptial tubercles
-
-
+
+
+
Ovaries
both
both
both
left
left
autumn spawning may return to fresh water the following spring
and are transparent and elongate; mostly mature adults return
one year later (age about 2 years) to spawn and die (see Jolly,
1967; McMillan, 1961; Milward, 1966).
Protolroclidae.-lAXXXe is known of this family, with one species
extinct the other rare. McDowall (1976) and Berra (1982) have
described what is known of life histories. The eggs are small
(~ 1 mm) round and demersal, and are probably deposited in
upstream fresh waters. The larvae are not known but believed
to be carried to estuaries or the sea to develop, probably for
about six months, and return to freshwater in spring as slender
transparent juveniles (Fig. 80).
Aplochilonidae. — \n Lovettia, mature adults migrate from the
sea in spring to spawn in fresh water, and are strongly dimorphic.
The male's reproductive opening migrates forward to the isth-
mus and the opercular flaps become elongated and papillated.
Fecundity is very low (=1 50-200). The tiny eggs (= I mm) are
demersal and spherical, and are attached in clusters to hard
surfaces (logs, stones, etc.) taking up to 23 days to hatch, and
the larvae drift downstream to sea. The post spawning adults
die. The life cycle is essentially annual. Larvae at hatching are
152
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 79. Young of Relropinna retropmna. 35 mm (above); and Aplochilon sp., 24 mm (below).
5-7 mm long with little yolk anteriorly below pectoral fins. They
are very elongate, the vent posterior at more than 75% body
length, the trunk encompassed by a low finfold from head around
tail to yolk sac. Small pedunculate pectoral fins occur. Pigmen-
tation is confined to the eyes and a narrow band in mid-ventral
between head and vent. Newly hatched larvae disperse to sea
and are not fiirther studied (see Blackburn, 1950).
Aplochilon taeniatus is recorded spawning in ft-esh water dur-
ing winter, the small (1.5 mm), spherical eggs being demersal
and attached to firm benthic objects, fecundity 2,500-3,000 and
development about 20 days. The larvae are very elongate and
slender with a yolk sac beneath the pectoral fins. The vent is at
about 75% of body length. A finfold encompasses the trunk and
tail. Campos (1969) shows a single large melanophore just in
front of the vent. His figure of a larva presumably 8 mm long
(he states 80 mm) shows a series of melanophores along the
abdomen and a few on the lower caudal peduncle. Recent col-
lections of larval Aplochiton from Fiordo Aisen in southern
Chile show that some movement to sea occurs. At a length of
24 mm the late larva has well differentiated rays in the dorsal,
caudal, anal and pectoral fins and distinct pelvic fin buds are
evident (Fig. 79). An adipose fin is also differentiated. Pigmen-
Fig. 80. Young of Galaxias maculatus. 14.5 mm (above); and Prolotrocles maraena 35 mm (below).
McDOWALL: SALMONIFORMS
153
tation is sparse, limited to spaced melanophores along the ab-
domen. The larva remams very elongate, the vent at about 85%
of total length. Eigenmann ( 1 928) reported that A. manmis (=A.
taenialus) spawns in the sea but this has never been corroborated
(see Campos. 1969).
Galaxiidae. — Diadromous species: Spawning is usually in fresh-
water. Eggs of Galaxiasfasciatus are deposited in autumn-win-
ter on stream-side forest debris during floods and develop out
of water, hatching when re-immersed during a subsequent flood.
The larvae go to sea on hatching, returning in spring as elongate,
transparent juveniles about 45 mm long. A minor metamor-
phosis involves shrinkage at freshwater entry. The eggs are of
moderate size (~2 mm) and number many thousands; devel-
opment takes about 30 days. Most other diadromous species
have unobserved habits. G. maciilatus spawns in tidal estuaries
where streamside vegetation is inundated at high spring tides
and development takes place between successive series of spring
tides. Most adults die after spawning and larval life is marine.
The eggs are simple, spherical, demersal and adhesive, varying
from 1-2 mm diameter and more or less colourless. Benzie
(1968a) described eggs of G. maculatus as "finely etched." Lar-
vae at hatching have a well developed yolk sac. with a single
oil globule, the sac below and behind the pectoral fins. The larvae
are slender and elongate at hatching, 7-8 mm long, and have
the finfold continuous from about mid dorsal around tail to yolk
sac. The vent is posterior, at about 75% of total length.
Non-diadromous species: Most species in the family are non-
diadromous (31 of 37 species). Those known spawn on sub-
strates near adult habitats and the pelagic "whitebait" juvenile
stage is omitted. Eggs are laid in aggregations (G. vulgaris). Lar-
vae on hatching, where described, resemble those of G. mac-
ulatus.
Galaxiella pusilla is distinctive in being sexually dimorphic,
spawning in pairs, the females laying eggs individually on stream
vegetation. Individual placement of eggs is also reported for
Brachygalaxias bullocki. The ability to aestivate is recorded for
some species (Neochanna. New Zealand) and spawning follows
restoration of water. It is suspected in others (Galaxiella. Aus-
tralia; Brachygalaxias. Chile) and may involve drought survival
of eggs (see Benzie, 1968a, b; Backhouse and Vanner, 1978;
Cadwallader, 1976; Campos, 1972; McDowall 1968b, 1978;
McDowall et al., 1975; Mitchell and Penlington, 1982).
Little IS known about the marine larval/juvenile life of any
of these southern salmoniforms. Small numbers of Galaxias
larvae (Fig. 80) have been collected at sea (McDowall et al.,
1975), as have a few, usually pre-migratory Retropinna. The
presence of a pelagic-living, transparent, elongate, migratory
juvenile seems to be common to most species that are marine
or lacustrine at some slage— Galaxias. Retropinna. Prototroctes.
Aplochiton. This is likely to have little phylogenetic significance
but to relate more to their pelagic, oceanic habits. These small
fish resemble many other unrelated fish with pelagic juveniles.
The marine, pelagic phase is followed in all instances by a minor
metamorphosis on entry to fresh water. Principally this involves
rapid assumption of pigmentation and in some species a distinct
change in body form. Shrinkage is recorded in a few species.
Identification of oceanic larvae and juveniles to family is
assisted by dorsal fin position and the early development of an
adipose fin in all but galaxiids. The elongate form with the vent
at about 75% of total length is helpful. Differences have been
recorded in pigment patterns between some of the diadromous
galaxiid juveniles although insufficiently to use as diagnostic
differences (McDowall and Eldon. 1980). Meristic differences
between species are of little value for specific identification ow-
ing to their wide ranges and latitudinal variability. Identification
remains a difficulty and improvement will depend on the capture
and examination of additional material.
Fisheries Research Division, Ministry of Agriculture and
Fisheries, Christchurch, New Zealand.
Osmeridae: Development and Relationships
M. E. Hearne
OSMERIDAE, the true smelts, are a small family of northern
hemisphere salmoniform fishes. The family includes 2
subfamilies, 6 genera, 10 species, and 13 forms (monotypic and
subspecies). They have marine, anadromous or landlocked and
freshwater life histories in the Pacific, Arctic and Atlantic oceans
and their drainages (McAllister, 1963). These silvery tasty little
fishes are captured by both recreational and commercial pur-
suits along the open coast beaches and rivers during their spawn-
ing runs.
Development
The smelts are highly selective spawners, choosing to spawn
on very specific sub-tidal areas, beaches and rivers. Some species
spawn in the daytime, and some spawn at night. The eggs of
osmerids possess an adhesive membrane that attaches to sand
grains and plant material. This anchor membrane results from
the ruptunng of an outer "chorion" during spawning, which
turns out and onto the substrate. This adaptation for demersal
spawning is observed in all 10 species of osmerids (Hamanda,
1961; Thompson et al., 1936; Morris, 1951; McAllister, 1963;
Simonsen, 1978; DeLacy and Batts, 1963; Hearne, 1983).
The first description of smelt development was made by Eh-
renbaum (1894) for the Elbe River smelt, Osmerus eperlans
illustrating embryological stages, yolk-sac larva, transforming
larva, and the juvenile. Up to now, the yolk-sac stage of many
of these species has been at least illustrated or photographed.
154
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 81. (A) Yolk-sac larvae of Spirinchus starksi. Osmeridae, 7.4 mm, from Morris (1951); (B) Yolk-sac larvae of Plecoglossus altivelis.
Plecoglossidae, ca. 6.0 mm, from Okada (1960); (C) Post yolk-sac larva of Salangichlhys inicrodon (Salangidae), ca. 7.0 mm, drawn from two
specimens in CAS 504 1 5.
Ahlstrom (pers. comm.) deteimined that, in general, osmerid
larvae were unique from other elongate larvae in the California
Current system by having a single mid-ventral row of mela-
nophores below the gut. Based on all the available larval de-
scriptions for osmerids, including the Atlantic forms, this single
row of melanophores appears to be a hallmark of the family.
Listed in Table 37 are sources of larval and juvenile descriptions
for the ten species of smelts.
These descriptions use various characteristics for each species
and are not comparative in design. Melanophore counts are
referred to by Yapchionges (1949), Follett (1952), Simonsen
(1978), Morris (1951), Dryfoos (1965) and Moulton (1970).
Myomere counts were used by Delacy and Batts ( 1 963). Cooper
(1978) and Morris (1951) used both myomere and melanophore
counts.
Larval osmerids have the following external features in com-
mon: elongate body shape; gut about 75% body length; mouth
sub-terminal; head dorso-ventrally flattened; lower jaw not well-
developed in early larvae; conspicuous choroid fissure in ventral
third of eye with ventral rim of clear choroid tissue; stalked
pectorals, stalk becoming more pronounced in late larvae; yolk
sac positioned 6-12 myomeres posterior to the pectoral base;
finfold extending from midbrain area to tail, from mid-yolk sac
to anus, and from anus to tail; no dorsal melanophores; scattered
melanophores (20-50) on ventral half of yolk sac; 0-2 mela-
nophores on posterior ventral half of yolk sac; single row of
melanophores along ventral midline of gut, sometimes extend-
ing into finfold; 1-3 melanophores on dorsal surface of gut at
the anal bend; single row of melanophores on ventral midline
of tail; conspicuous opaque liver ventral to foregut (Ehrenbaum,
1894; Yapchionges, 1949; Morris, 1 95 l;DeLacy and Batts 1963;
Dryfoos, 1965; Eldridge, 1970; Blackburn, 1973; Cooper, 1978;
Heame, 1983).
A comparative study of four of the species oflTOregon (Heame,
1983) used ventral melanophore counts and myomere counts
in an attempt to characterize the larvae of these species. Ten-
dencies in these counts showed Hypomesus pretiosus and Spi-
rinchus starksi to have high ventral melanophore counts while
Spirinchus thaleichthys and Thaleichthys pacificus have lower
melanophore counts. Myomere counts showed tendencies that
further separated each similarly pigmented pair.
Table 37, Sources of Larval and Juvenile Descriptions of Smelts, (x
no description found.)
Taxon Larvae Juveniles
Hypomesus
pretiosus
Hypomesus
transpacificus
Spirmchus
lanceolatus
Spirinchus
starksi
Spirinchus
thaleichthys
Thaleichthys
pacificus
Alios merus
elongatus
Mallotus villosus
Osmerus mordax
Osmerus eperlanus
Yapchionges, 1949
X
Hikita, 1958
Morris, 1951
Dryfoos, 1965;
Moulton, 1970
DeLacy and Batts,
1963
X
Schmidt, 1906c
Cooper, 1978
Ehrenbaum, 1894
Follett, 1952
Simonsen, 1978
Hikita, 1958
Heame, 1983
Simonsen, 1978
Baraclough, 1964
Heame, 1983
Templeman, 1948
Cooper, 1978
Ehrenbaum, 1894
HEARNE: OSMERIDAE
155
The transformational stages of osmerids are not fully known,
since complete developmental series have not been reported for
all of the species. However, it is apparent from rearing studies
(Morris, 1951; Cooper, 1978) that caudal flexion occurs after
yolk absorption and along with median fin formation. The pelvic
fins arise from the ventral body musculature as prominent buds
after the median fin rays have formed, and appear stalked, be-
coming inserted as the ventral musculature joins ventrally. The
pectoral fins are present at hatching and remain pedunculate
until postflexion stages acquire adult-like pigmentation.
During flexion an additional series of melanophores forms
along the ventro-lateral edge of the body musculature and ap-
pears as a double row of spots from ventral view. There are also
count differences between the species in these secondary me-
lanophores (Heame, 1983), and they may aid in identification
of flexion and postflexion stages.
The postflexion stages of two species of osmerids have been
erroneously described as new species belonging to other families
by Chapman ( 1 939). Hubbs (1951) has shown that one of these
smelts, placed in the family Paralepididae as Lestidium parn.
is actually a late postflexion stage of Thaleichthys pacificus. and
the other one, placed in the family Sudidae as Sudis squamosa,
is a postflexion Mallotiis villosus. The blackened gut cavities of
the postflexion stages of these two species, lend a distinct re-
semblance to the midwater-inhabiting sudids and paralepidids,
and also suggest a unique departure from the developmental
trend of the other species that may warrant the use of the term
"pre-juvenile" as defined by Hubbs (1943).
Relationships
In a recent statement on classification, Rosen ( 1974) proposed
an infraorder Salmonae to include two suborders, the Argen-
tinoidei and Salmonoidei, the Osmeridae being placed in the
latter under the superfamily Osmeroidea (with the Plecoglos-
sidae, Retropinnidae, and Salangidae). On the basis of embry-
ological and larval features, Soin (1980) characterized different
types of salmoniform fishes. He placed the Piecoglossidae and
Osmeridae in the same category based on similar egg mor-
phology (presence of an anchor membrane), degree of devel-
opment at time of hatching and at time of yolk absorption. In
a study of stomiiform fishes using adult characters. Fink and
Weitzman (1982) placed the families Osmeridae, Salangidae,
Piecoglossidae, Retropinnidae, and Galaxiidae all together as
"unresolved sister taxa."
The larvae of osmehds (Spin nchus slarksi. Fig. 8 1 A) are strik-
ingly similar to larval plecoglossids (Plecoglossus alttvelis. Fig.
8 1 B). The yolk sac of these two families is positioned such that
its posterior edge is near myomere 11-12. The plecoglossids
also have a single median ventral row of melanophores and, as
development proceeds, another latero-ventral row of spots ap-
pears along the ventral edge of the body musculature, just as in
osmerid development.
Photographs of the yolk-sac stage of Salangichthys microdon,
Salangidae, (Okada, 1960: pi. 17) show that the yolk-sac mor-
phology is different than in the Osmeridae and Piecoglossidae.
The yolk sac of Salangtchthys microdon is co-extensive with the
undersurface of the gut and is more oblong shaped (pyriform)
than the more rounded, anteriorly placed yolk sac of the os-
merids and plecoglossids. The post yolk-sac larvae of salangids
(Fig. 81C) are nearly identical to those of osmerids and pleco-
glossids, exhibiting the single median ventral row of melano-
phores. Also, the eggs of salangids are different than the osmerid-
plecoglossid type by having, instead of an anchor membrane,
an anchoring structure that is composed of various kinds of
filaments that turn out and onto the substrate (Wakiya and
Takahashi, 1913). Larval development is not yet documented
for the Sundasalangidae, however adults of this minute family
of salangoid fishes have ventral pigment patterns (Roberts 1981:
fig. 1) that are strikingly similar to the postflexion pigment pat-
terns of osmerids. The same ventral pigment patterns (single
ventral midline, paired latero-ventral melanophores) can also
be seen in adults of Salangidae (Okada, 1960).
One interpretation may be that the similarities in ventral
pigment patterns and egg morphology may be the retention of
a trait of an ancestor common to the Osmeridae, Piecoglossidae,
and Salangidae, and give support to theories arising from sys-
tematic observations of adult salmonoids that these families are
closely related to each other and not to the other salmoniform
families.
184 Day Street, San Francisco, California 94131.
Argentinoidei: Development and Relationships
E. H. Ahlstrom, H. G. Moser and D. M. Cohen
THE argentinoid fishes as here discussed have been consid-
ered a suborder by Cohen ( 1 964b) and many other authors
and a super-family of an expanded suborder that also includes
the alepocephaloids by Greenwood and Rosen (1971). The latter
group is not treated at length in this book, because little infor-
mation on alepocephaloid ELH stages has appeared since Beebe's
(1933a) survey which showed they hatch from large eggs and
have direct development. The argentinoids sensu strictu appear
to be monophyletic on the basis of four derived characters. One
character concerns the development of rays in the finfold of the
larva and is described later in this paper. A second character is
the development of pustules on the inner surface of the chorion
(not known for opisthoproctids). A third character relates to the
swimbladder, which, when present, is served by a unique kind
of rete mirabile, first described by Fange (1958) and further
investigated by Marshall (1960) who named these structures
micro-retia mirabilia. A fourth unique character, and one which
never has been adequately studied and documented, is the ten-
dency in the group for the vomer and palatines to assume the
functions of the premaxillary and maxillary.
156
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 38. Literature References for Ontogenetic Stages of Argentinoids.
Species
Egg
Transformation stage
Argentinidae
Argentina elongata
A. silus
A. sphyraena
Glossanodon leioglossus
G. polli
G. semifascialus
Microstoma microstoma
Nansenia groenlandica
N. oblita
Xenopthahnichthys danae
Bathylagidae
Balhylagus antarcticus
B. euryops
B. longirostris
B. nigrigenys
B. ochotensis
B. schmidti
B. stilbius
B. wesethi
Opisthoproctidae
Balhylychnops exitis
Dolichopteryx spp.
Dolichopteryx longipes
Macropinna microstoma
Opisthoproctus grimaldii
Rhynchohyalus natalensis
H'interia telescopa
Robertson, 1975a
Schmidt, 1906c
Sanzo, 193 Id
Sanzo, 193 Id
Nishimura, 1966
Sanzo, 193 Id
Sanzo, 193 Id
Yefremenko, 1982
Pertseva-Ostroumova
and Rass, 1973
Ahlstrom, 1969
Ahlstrom, 1969
Ahlstrom, 1969
Holt, 1898; Schmidt, 1906c
Schmidt, 1906c, Sanzo, 193 Id
Schmidt, 1918, Sanzo. 193 Id
Nishimura, 1966
Lo Bianco, 1903; Schmidt, 19 If
Sanzo, 193 Id
Schmidt, 1918
Schmidt, 1918; Sanzo, 193 Id
Yefremenko, 1979a, 1983
Brauer, 1906; TSmng, 1931
Ehrenbaum, 1905-09; Murray
and Hjort. 1912; Roule and
Angel, 1930; Beebe, 1933b
Pertseva-Ostroumova and Rass,
1973
Ahlstrom, 1972b
Dunn, 1983a
Ahlstrom, 1965, 1972b
Ahlstrom, 1965, 1972b
Roule and Angel, 1930
Beebe, 1933a
Chapman, 1939
Schmidt, 1918
Bertelsen et al., 1965
Belyanina, 1982b
Schmidt. 1906
Sanzo, 1931d
Poll, 1953
Nishimura, 1966
Schmidt, 1918
Schmidt, 1918
Schmidt, 1918
Bertelsen, 1958
TSning, 1931
Beebe, 1933b
Ahlstrom, 1972b
Dunn, 1983a
Cohen, 1960
Although now there seems to be general agreement as to the
genera to be included in the group, their internal arrangement
is an unsettled matter. Opinions range from those of C. L. Hubbs
(1953), who relegated all to a single family, to those of Chapman
(1948 and papers cited therein), who advocated eight different
families. Subsequently Cohen (1964b) classified the group in
three families using inadequately evaluated characters.
Family Argentinidae (most genera are probably worldwide):
Subfamily Argentininae (benthopelagic. outer shelf to slope):
Argentina (12 species) and
Glossanodon (seven or more species).
Subfamily Microstomatinae (mesopelagic)':
Microstoma (one or two species),
Nansenia ( 1 3 species) and
Xenophthalmichthys (one or two species).
Family Bathylagidae (meso-to bathypelagic):
Bathylagns (including Leiiroglossiis and Therobromus; about
a dozen to 1 5 species; several species in the Arctic and
Antarctic).
Family Opisthoproctidae (mesopelagic):
Group 1:
Macropinna (one species; restricted to N. Pacific and east-
em S. Pacific),
' Herein considered a distinct family.
Opisthoproctus (two species),
Rhynchohyalus (one species; Atlantic and Indian Oceans)
and H'interia (one species).
Group II:
Balhylychnops (one or more species), and
Dolichopteryx (perhaps half a dozen species).
An alternate arrangement presented by Greenwood and Ro-
sen (1971) and essentially based on inadequately evaluated char-
acters in the branchial arches and caudal fin skeleton proposed
two families within a superfamily Argentinoidea: Family Ar-
gentinidae and Family Bathylagidae with Subfamily Bathyla-
ginae (including Microstomatidae) and Subfamily Opistho-
proctinae.
Unanswered questions concerning the systematics of the group
are numerous and exist at all levels. Following is a summary.
( 1 ) What are the external relationships of the argentinoids? (2)
How many distinct lineages exist within the group, how should
they best be arranged with respect to each other, and how many
families should be recognized? (3) Do Argentina and Glossan-
odon constitute a monophyletic group? If not, where does each
belong? (4) How many genera should be recognized among the
bathylagids? (5) Within the opisthoproct group do the elongate
species in the Bathylychnops-Dolichopieryx group and the short-
bodied species in the Opisthoproctus group constitute mono-
phyletic lineages and if so should they be named? (6) Since
species complements of genera are inadequately known, espe-
AHLSTROM ET AL.: ARGENTINOIDEI
157
Table 39. Characters of the Eggs of Argentinoidei.
Number of
Distribution of
Diameter of
Species
Diameter
oil globules
oil globules
oil globules
Source
Argentina stalls
1.31-1.66
vegetal pole
0.27-0.46
Original
Argentina siliis
3.0-3.5
vegetal pole
0.95-1.16
Schmidt, 1906c
Argentina sphyraena
(Mediterranean)
1.60-1.68
vegetal pole
0.44
Sanzo, 193 Id
(North Sea)
1.70-1.85
vegetal pole
0.37-0.47
Schmidt, 1906c
Argentina elongata
1.67-1.80
vegetal pole
0.35-0.45
Robertson, 1975a
Glossanodon leioglossus
1.44-1.52
vegetal pole
0.36
Sanzo, 193 Id
Glossanodon semifasciatus
1.5-1.6
vegetal pole
0.36
Nishimura, 1966
Microstoma microstoma
(Atlantic)
1.60-1.72
vegetal pole
0.48-0.52
Sanzo, 193 Id
(Pacific)
2.05-2.38
vegetal pole
0.49-0.82
Original
Nansenia Candida
1.39-1.56
vegetal pole
0.41-0.49
Original
Nansenia crassa
1.05-1.30
vegetal pole
0.30-0.35
Original
Nansenia ohlita
1.39-1.56
vegetal pole
0.40-0.53
Sanzo, 193 Id
Bathylagiis antarclicus
1.8-2.2
3-8
*
0.2-0.3
Yefremenko, 1982
Bathylagus schmidti
1.65-1.90
up to 9
*
Ahlstrom, 1969
Bathylagiis slilhiiis
1.01-1.21
15-25
*■
Ahlstrom, 1969
Bathylagus urotranus
1.03-1.21
15-25
*
Pertseva-Ostroumova and Rass,
1973, and original
Bathylagus ochotensis
0.92-1.1
many to
two clumps
**
Original
Balhylagits wesclhi
0.90-1.10
12-20
**
Ahlstrom, 1969
Bathylagus nigrigenys
0.83-1.09
12-20
**
Pertseva-Ostroumova and Rass,
1973, and original
' First grouped at vegetal pole, then move to beneath embryo, then coalesce to one at each equatonal pole.
• Numerous globules at vegetal pole then coalesce to one clump at each equatonal pole.
cially the mesopelagic ones, do presently available early life
history specimens help define the species composition of argen-
tinoid genera?
Development
Eggs are known for 1 3 species of argentinoids and larvae for
22 species (Table 38). We present in this paper eggs of 5 ad-
ditional argentinoid species and larvae of 8 additional species.
These are: eggs and larvae oi Argentina sialis. Microstoma sp.,
Nansenia Candida and N. crassa; larvae only for Bathylagus
argyrogaster. B. bencoides. B. pacificus and Balhylychnops ex-
ilis: eggs only for Bathylagus ochotensis.
Eggs
The eggs of argentinoids are pelagic, round, have a moderate
to narrow perivitelline space, segmented yolk and a chorion
with distinctive pustules on the inner surface (Table 39, Fig.
82). Egg diameters and oil globule characters are given in Table
39.
Argentinoid larvae hatch as relatively undifferentiated yolk-
sac larvae, regardless of egg size. That is, yolk-sac larvae of A.
silus at 7.5 mm, newly hatched from eggs 3.0-3.5 mm diameter,
are at about the same stage of development as 3 mm bathylagid
yolk-sac larvae which hatch from 1 mm eggs. In most marine
fishes larger eggs produce more highly differentiated hatchlings.
Larvae
Body form. — Argentinid and bathylagid larvae are slender, those
of microstomatids are deeper-bodied, and opisthoproctids have
a wide variety of body shapes ranging from the slender larvae
of Balhylychnops to the deep-bodied Opisthoproctus (Table 40,
Figs. 83-87).
The gut is elongate and straight in argentinids and bathylagids,
with the exception of B. milleri where the gut is straight but
only about half the body length. In argentinids the gut is lined
with transverse rugae for almost the entire length. In most bath-
ylagids the gut has two distinct sections: an anterior section with
longitudinal internal ridges, separated by a valve from a shorter
posterior section with transverse rugae. The anterior section in
B. hericoides and B. longirosths is markedly smaller in diameter
compared with other species. Larvae of j5. wesethi. B. nigrigenys
and B. argyrogasterhave transverse rugae along the entire length
of the gut and the anterior section is relatively larger in diameter
and thin-walled. Also the posterior section is subdivided by a
second valve. B. ochotensis larvae develop a similar structure.
The gut in microstomatid larvae is long, but anteriorly has
an elongate S-shaped fold that lies flat on the left side (Fig. 84).
The lumen of the anterior folded section is characterized by
longitudinal ridges whereas the posterior straight section has
transverse rugae. The short pyloric section has longitudinal ridges.
Schmidt (1918) shows the gut extended beyond the finfold mar-
gin in Nansenia ohlita and trailing in early stage Microstoma
microstoma larvae but we have not seen this in any specimens
of these genera.
In opisthoproctids the gut is elongate in Balhylychnops and
Dolichoptery.x and relatively shorter in the deeper-bodied gen-
era, Macropinna. Rhyncholyalus and Opisthoproctus. In all gen-
era there is a sac-like stomach, which exits through a constricted
pyloric section to the intestine. In Balhylychnops and Doll-
chopteryx the sac is elongate and pointed at its tip whereas in
the other genera it is more rounded in form. The sac lies on the
left side, except in Balhylychnops where it lies on the right. In
the latter genus the pyloric constriction leads into a short but
prominent bulbous section. DoHchopteryx is similar but lacks
158
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig, 82. Eggs of argentinoids. (A) Argentina sialus. 1.5 mm, CalCOFI 5103, Sta. 1 17.35; (B) Microstoma sp., 2.2 mm, CalCOFI 751 1, Sta.
87.90; (C) Nansenia Candida. 1.4 mm, CalCOFI Sta. 60.90; (D) N. crassa. 1.5 mm, CalCOFI; (E) Bathylagus stilhius. 1.1 mm, from Ahlstrom
(1969); (F) B. schmidli. 1.8 mm. from Ahlstrom (1969); (G) B. ochotensis. 1.1 mm, CalCOFI 5002 Sta. 60.90; (H) B. weselhi. 1.0 mm, Ahlstrom
(1969); (I) B. nigrigenys. 0.96 mm. CalCOFI 5106 Sta. 157.20.
the post-pyloric bulb. In Macropinna and Opisthoproctus there
is a straight section leading posteriorly from the pylorus, which
ends in an S-shaped fold and an enlarged rectal bulb, the latter
described by Bertelsen and Munk (1964). The anterior section
including the sac and pyloiojs have longitudinal internal ridges
while sections posterior to this have transverse rugae. In late
larval stages the entire section posterior to the pylorus becomes
part of the S-shaped coil.
The head is relatively small in argentinids and has a rounded
blunted anterior profile (Fig. 83, Table 40). It is slightly larger
in most microstomatids, with the exception oi Microstoma sp.
(Pacific form) which has a small head. In most microstomatids
the head has a rounded, blunted anterior profile and is bent
slightly downward from the longitudinal axis. In both families
the eye is either round or slightly ellipsoidal. In bathylagids the
head is moderate in size but highly various in shape (Figs. 85,
86; Table 40). The snout is generally longer than in Argentinidae
and Microstomatidae.
Eye shape and structure vary greatly within the bathylagids.
Bathylagus milleri has a large, nearly round eye in contrast to
AHLSTROM ET AL.: ARGENTINOIDEI
159
v^^&^ t&. **J isf *tii "jv tn^ 'zr
• •^^^I''' ."»• *. *S ^^ S -S^*
» ^i^^-^j*** « #»♦ "^ »? PIt^i
Fig. 83. Larvae of Argentinidae. (A) Argentina stalls. 7.0 mm. CalCOFI 5103 Sta. 1 17.35; (B) A. stalls. 9.0 mm, CalCOH 5104 Sta. 97.40;
(C) .-1. stalls. 17.5 mm, CalCOFI 5103 Sta. 120.35; (D) A. stalls. 21.0 mm, CalCOFI 5105 Sta. 123.40; (E) A. silus. 32.5 mm. redrawn from
Schmidt (1906c); (F) A. sphyraena. 19.2 mm, ibid; (G) Glossanodon semifasciatus, 12.5 mm, from Nishimura (1966).
AHLSTROM ET AL.: ARGENTINOIDEI
161
Table 40. Comparative Morphometry of Aroentinoid Larvae. Mean values (%) of body proportions for three ontogenetic stages (preflexion-
flexion-postflexion) are listed.
Snoul-anus
Eye stalk
Snout-anal
Snout-dorsal
Snout-pelvic
distance
Head length
Head width
Eye length
length
Body depth
fin distance
fin distance
fin distance
Species
Body length
Body length
Head length
Head length
Head length
Body length
Body length
Body length
Body length
A rgentma sialis
76-78-84
17-21-22
54-44-41
28-24-24
—
9-10-10
0-78-81
0-46-47
0-0-49
Microstoma microstoma
?-?-80
7-7-23
7-7-45
7-7-27
—
7-7-13
7-7-80
7-7-68
0-7-64
Microstoma sp. (Pacific)
76-79-80
17-19-19
53-49-44
31-29-27
_
8-10-10
0-78-81
0-70-72
0-64-67
Nansenia Candida
74-77-82
21-25-26
60-50-44
36-28-28
—
12-14-16
0-75-82
0-54-58
0-56-61
Nansenia crassa
74-78-80
22-25-28
58-50-44
36-29-24
_
10-12-15
0-76-80
0-52-57
0-56-60
Nansenia groenlandica
7-78-80
7-27-25
7-50-42
7-21-23
—
7-15-15
7-77-80
7-52-52
7-54-57
Xenophthatmichthys danae
?-?-82
7-7-24
7-7-48
7-7-21
—
7-7-12
7-7-86
7-7-74
''-7-52
Bathylagus milleri
59-57-61
20-19-26
56-54-52
31-27-26
_
9-9-15
0-0-71
0-0-50
0-0-45
Bathylagus schmidli
72-76-78
16-19-22
50-52-46
39-26-25
.04-0-0
7-8-10
0-0-79
0-0-57
0-0-55
Bathylagus slilbius
74-77-80
20-22-24
54-53-47
32-25-20
.03-0-0
8-10-13
0-0-79
0-0-57
0-0-55
Bathylagus urotranus
78-82-81
20-24-28
56-53-46
27-18-21
.03-0-0
10-10-12
0-0-81
0-0-61
0-0-59
Bathylagtis pacificus
76-85-81
22-24-25
39-42-44
29-22-18
28-29-20
8-10-13
0-81-80
0-49-48
0-51-51
Bathylagus curyops
78-80-82
18-20-20
46-50-50
31-26-25
10-7-3
10-11-12
0-78-80
0-45-48
0-0-47
Bathylagus bericoides
84-85-89
25-26-26
34-38-36
27-25-22
60-64-36
8-8-9
0-83-88
0-0-52
0-0-53
Bathylagus longiroslris
85-88-92
26-27-25
34-34-34
24-20-19
54-48-27
8-10-10
0-88-90
0-0-53
0-0-57
Bathylagus ochotensis
81-85-90
20-23-23
44-44-44
32-21-21
17-15-15
8-10-11
0-83-87
0-53-54
0-56-56
Bathylagus wesethi
79-89-94
13-26-27
59-53-50
27-16-13
—
9-14-16
0-85-90
0-58-60
0-57-59
Bathylagus nigrigenys
80-86-93
20-29-28
78-60-53
30-18-14
—
12-16-18
0-86-90
0-57-60
0-0-60
Bathylychnops exilis
7-80-82
7-21-22
7-42-38
7-22-18
—
7-8-7
7-82-84
7-71-73
7-66-67
Dolichopleryx longipes
7-74-75
7-24-26
7-44-34
7-22-16
—
7-8-10
7-0-77
7-0-71
7-62-62
Macropinna microstoma
7-64-59
7-26-35
7-52-47
7-22-21
—
7-15-21
7-0-70
7-0-66
7-43-48
Opisthoproctus soleatus
7-7-80
7-7-37
7-7-46
7-7-18
-
7-7-18
7-7-83
7-7-63
7-7-40
other species which have relatively smaller, more elliptical eyes.
Eyes are sessile in B. milleri and in the B. wesetht group but are
stalked to some degree in all other species known. In B. slilbius
and relatives (B. urotranus, and B. schmidli') the stalks are short
and found only in early larvae. Stalks are longer and persist into
later larval stages in other species, reaching a ma.ximum of 65%
of the head length in B. bericoides.
In opisthoproctids the head is moderate in size in the slender
forms, Bathylychnops and Dolichopleryx, and longer and more
massive, with a pronounced hump or bend at the nape, in the
deep-bodied genera. All genera have an elongate snout and Bath-
ylychnops has a unique triangular flap at its tip. Bathylychnops
has round eyes that are rotated slightly dorsoanteriad. In the
other genera, the eyes are tubular and directed dorsally, even
in the smallest larvae available. Eye diverticulae with associated
accessory retinae, characteristic of opisthoproctid adults, begin
to form at the end of the larval period.
Fins —A major feature of all argentinoid larvae is the devel-
opment of a prominent median finfold in which the dorsal and
anal fins develop, connected to the trunk by a series of hyaline
strands (Figs. 83-87). The first fins to form are the pectorals. In
argentinids and bathylagids they are relatively small and de-
velop rays late in the larval period. Microstomatid and opis-
thoproctid pectoral fins are generally larger; however, there is a
wide size range, from relatively small fins in Microstoma to
large, fan-like fins in some species of Nansenia (e.g., N. groen-
landica) to very elongate pectorals in Dolichopleryx binocularis.
Ossification of rays begins earlier in these groups, usually before
notochord flexion.
After the pectorals, the caudal fin is usually the next to form.
In argentinids notochord flexion and development of principal
caudal rays occurs at a size about midway in larval growth
whereas in opisthoproctids this occurs earlier in the larval pe-
riod. In bathylagids the process is somewhat delayed and in
some species (e.g., B. euryops. B. milleri) notochord flexion may
not be completed until near the end of the larval period.
The dorsal and anal fins begin to form at about the stage of
notochord flexion in all argentinoids except opisthoproctids,
where notochord flexion slightly precedes the appearance of
dorsal and anal fins. The anal fin begins forming far posteriad
in argentinoids, just posterior to the anus or the point of de-
flection of the free terminal gut section. In B. milleri and in the
deep-bodied opisthoproctids with coiled guts there is a space
between the anus and the anal fin origin.
The position of the dorsal fin is varied among argentinoids
and forms in the larvae in approximately the same position that
it will occupy in the adult. The fin has its most anteriad location
in Argentina where its origin is well forward of the midpoint of
the body (Fig. 83). The extreme case is found in A. silus where
snout to dorsal origin is about 38% of the body length in larvae
and about 43% in adults. In most bathylagids the dorsal origin
is slightly anterior to mid-body. The exceptions are B. slilbius
and relatives, where the dorsal origin is slightly posterior to
mid-body, and B. wesethi and relatives where it is located still
further posteriad.
Fig. 84. Larvae of Microstomatidae. (A) Microstoma microstoma. 1 1.0 mm, from Schmidt (1918); (B) Microstoma sp., 12.0 mm, CalCOFl
5 104 Sta. 90.52; (C) Nansenia Candida. 8.4 mm, CalCOFl 5007 Sta. 1 00.70; (D) N. crassa. 8.5 mm, CalCGR 5 103 Sta. 1 37.50; (E) N. groenlandica,
10.0 mm, from Schmidt (1918); (F) N. oblita, 9.0 mm, ibid; (G) Xenopthalmichthys danae. 16.5 mm, from Bertelsen (1958).
162
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 85. Larvae of Balhylagus. (A) B. milleri. 27.5 mm, CalCOFl 5106 Sta. 70.60, dorsal view of 9.5 mm specimen at left; (B) B. schmidli.
31.5 mm, CalCOFI Northern Holiday Exped. Sta. 31; (C) B. snlbiits, 23.2 mm, CalCOH 4905 Sta. 1 1 1.38, dorsal view of 8.5 mm specimen at
left;(D) B. pacificus. 21.4 mm, CalCOFI 7905 Sta. 63.60; (E) B. euryops 24.0 mm, dorsal view of 14.0 mm specimen at left, from Tuning (1931);
(F) B. antarcticus. 26.5 mm, from Yefremenko (1983).
AHLSTROM ET AL.: ARGENTINOIDEI
Table 41. Meristics of Argentinoid Fishes.
163
Branchiostegal
Dorsal
Anal
Pectoral
Pelvic
Procurrenl
Species
Venebrae
rays
fin rays
fin rays
lin rays
fin rays
caudal fin rays
Argentina
altceae
43-46
5
11-13
13-15
16-18
10-12
australiae
50-53
5
10-12
12-13
13-14
11-13
hrucei
45-47
5
10-12
11-13
18-20
13-14
elongata
52-55
5
10-12
11-14
13-16
11-12
euchus
48-49
5
12
13-15
16-18
10-11
georgei
48-51
5
10-12
10-13
16-19
12-14
kagoshimae
51-52
5
10-12
11-13
15-17
11-12
sialis
47-51
5
10-13
12-15
11-18
10-12
12-Hl
silus
65-70
6
11-13
11-17
15-18
12-13
sphyraena
46-55
6
10-12
11-15
12-15
10-12
stewarti
53-54
5
10-12
12-13
18-21
13-15
striata
48-52
5
10-12
11-14
18-21
11-15
10-1-9
Glossanodon
leiglossus
49-51
5
12-14
10-13
19-22
11-12
tmeatus
4
11-13
15
18-21
11-13
mildredae
50-52
5
13
13
23
12-13
polli
5
12-14
11-14
19-22
12-13
pygmaeus
43-44
5
10-12
11-13
12-14
10-12
semifasciatus
49
5
11-13
11-13
18-21
10-12
struhsakeri
51-53
12-14
12-13
23-25
13-15
Microstoma
microstoma
45-47
3-4
11-12
8-9
8
9-11
11 + 11
sp. (Pacific)
49-50
4
9-11
7-8
11
9
10-
-lH-10
Xenophthalmichthys
danae
3
10-12
9-10
7
8-9
10-1-9
Nansenia
atlanlica
41-42
4
9-10
8-9
12-13
10-11
ardesiaca
46-48
4
9-10
9-10
11-14
10-12
Candida
44-47
3
9-10
8-9
9-11
9-11
11-1-14
crassa
43-46
4
9-10
8-9
11-13
10-11
groenlandica
42-45
3
9-10
8-10
11-13
10-12
ohiita
42-45
4
10-11
9-10
10-11
10-11
Bathylagus
amarcticus
2
9-11
21-25
9-10
argyrogasler
2
12
14-15
8
bericoides
48-53
-)
10-11
18-22
10-12
9-10
euryops
44-46
2
9-11
16-19
7-12
7-9
greyae
2
11-13
13
12-13
10-11
longirostris
48-51
2
10-12
19-21
9-12
9-10
mtlleri
51-55
2
6-9
20-28
11-16
6-8
16-
-18-1-15-17
nigrigenys
41
2
11-12
14-17
10
8-10
ochotensis
47-49
2
9-12
12-15
9-11
9-10
13-
-14+15-16
pacificus
45-49
2
8-9
15-22
7-11
7-10
13+13-14
schniidti
47-52
2
10-11
11-14
8-9
8-9
16-
-17+16
stilhius
38-42
2
9-11
11-14
8-11
8-10
12-
-16+13-15
urolranus
39-42
2
9-10
10-11
9-11
7-8
12-
-14+12-13
weselhi
43-46
2
12-13
14-16
10-11
9-11
14-15-1-14-15
Dolichopteryx
anascopa
2
10
12
14
12
bmocularis
2
15
11
14
9
hrachyrhynchus
2
13
12
13
8
longipes
41-44
2
10-11
8-9
13
8-9
Bathylychnops
exilis
81-84
2
14-16
13-14
12-13
7
RhynJichyalus
natalensis
4
10-12
10
19-20
11-12
Macropinna
microstoma
36
3
11-12
14
17-19
10
Winteria
telescopa
8
8
12-14
9
164
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 41.
Continued.
Species
Vertebrae
Branchios
rays
legal
Dorsal
fin rays
Anal
tin rays
Pectoral
fin rays
Pelvic
fin rays
Procurrent
caudal fin rays
Opisthoproctus
grimaldii
soleatus
31
2
2
12-14
10-12
8
13-14
11
9-11
10
The dorsal fin forms in a variety of positions among micro-
stomatids. In most species of Nansenia. the dorsal fin originates
slightly posterior to mid-body, although in some species (e.g.,
A', groenlandica), its origin is slightly anterior to mid-body. The
dorsal origin is further posteriad in Microstoma. In M. microsto-
ma predorsal length is about 67-68% of the body length and
assumes a more anterior position in adults (ca. 63%). In larvae
of the Pacific species predorsal length is about 75% of the body
length, and is slightly more posteriad in adults. In adult Xen-
ophthalijuchthys the dorsal origin is at mid-body; however, in
the 16.5 mm specimen from the Atlantic (Bertelsen, 1958) pre-
dorsal length is 62% of the body length. In our single larva (12.2
mm) from the Pacific predorsal length is 75% of body length,
indicating a marked anteriad migration during ontogeny or strong
allometric growth posterior to the dorsal fin. Alternatively, the
Pacific form may prove to be distinct when adult specimens are
captured.
The dorsal fin in opisthoproctids is located posteriad on the
body. This is most marked in the slender forms, Bathylychnops
and Dolichopteryx. and reaches an extreme in D. hinocularis
where predorsal length is greater than Vj of the body length. In
the deep-bodied genera the dorsal origin is posterior to mid-
body, but less so than in the slender-bodied forms.
The pelvic fins are the last fins to form in most argentinoids,
usually late in the larval period. The exception is opisthoproctids
where the pelvic fins form early in the larval period. In argen-
tinids, bathylagids and microstomatids the pelvic fins form at
about mid-body, below the dorsal fin. In the slender opistho-
proctid genera the pelvics form well back on the body, but
anterior to the dorsal fin. Among the deep-bodied genera, Op-
isthoproctus forms the pelvics far back on the body, beneath the
dorsal fin. In Rhynchohyalus and Macropinna the pelvics de-
velop just posterior to mid-body and anterior to the dorsal fin.
In the larvae the fins are elevated to the sides of the body. This
position persists in juvenile and adult Macropinna where the
fins are located just behind and below the pectoral fin bases.
The pelvic fins become elongate in Dolichopteryx and the deep-
bodied genera. The pelvic fin base is pedunculate in opistho-
proctid larvae, a condition that persists into the adults of some
genera, notably Dolichopteryx. Argentinoids, except Microsto-
ma. Xenophthalmichthys and some species of Dolichopteryx.
develop adipose fins late in the larval period.
A summary of meristics of argentinoids is given in Table 41.
The sequence of ossification of fins and other skeletal elements
o( Bathylagus schmidti is described by Dunn (1983a).
Pigmentation. — \n argentinids, pigmentation consists of a series
of 6-8 ventral trunk blotches that extend from the pectoral fin
base to the end of the gut (Fig. 83). The series is continued
posteriorly as I or 2 median ventral blotches and ends as a large
blotch at the caudal region. The number of blotches is constant
for each species, as is the sequence of formation. In Argentina
sialis and Glossanodon the ventral blotches expand dorsally as
lateral bars, but this does not occur in A. silus and A. sphyraena.
These latter species differ additionally in lacking the internal
head pigment which develops in A. sialis and Glossanodon lar-
vae.
A feature common to most microstomatid larvae is a heavy
line of embedded pigment above the gut (Fig. 84). In Micro-
stoma this pigment continues forward to the gill arches and
within the head anteriorly to the snout. In Nansenia, head pig-
mentation is superficial, or concentrated ventrally on the head.
In Microstoma, an embedded dorsal line of pigment is located
posterior to the dorsal fin. Dorsal pigmentation in Nansenia
may take the form of a series of embedded blotches (e.g., N.
crassa) or an embedded line of melanophores running the length
of the body (e.g., N. ohlita). Most microstomatids have con-
spicuous melanistic pigment associated with the caudal fin re-
gion. A notable feature oi Microstoma and some Nansenia (e.g.,
N. crassa) is the presence of heavy melanistic pigment at the
curve of the gut loop. Our single damaged specimen of Xenoph-
thalmichthys (12.2 mm) has pigmentation similar to Micro-
stoma but lacks the posterior dorsal body pigment and has a
series of slanted melanophores along the hypaxial myosepta.
Pigment patterns in bathylagids may be grouped into two
categories— those species with large isolated melanophores (Fig.
85) and those with linear series of smaller melanophores (Fig.
86). Bathylagus milleri has a unique pattern of opposing dorsal
and ventral midline melanophores, large melanophores on the
head and pectoral fin base and a large lateral blotch on the
notochord tip.
Bathylagus stilbius and B. urotranus develop a series of 5-6
melanophores on each side of the posterior section of the gut.
A single large melanophore, is found on the lower trunk midway
between the pectoral fin and the anus and the head has mela-
nophores, chiefly on the upper and lower jaws and opercle (Fig.
85). B. schmidti differs in having a series of lower trunk blotches
and 1 or 2 postanal lateral blotches.
Bathylagus euryops has a series of 3-6 melanophores on the
lateral surface of the gut and 3-5 large melanophores on the
lateral surface of the trunk (Fig. 85). Other pigmentation consists
Fig. 86. Larvae of Bathylagus. (A) B. hericoides. \1 .1 mm, Dana Sta. 4007, dorsal view ofl 1.8 mm specimen at left; (B) B. longirostris. 20.1
mm, SIO/STOW XIII Exped., dorsal view of 12.4 mm specimen at left; (C) B. ocholensis. 21.5 mm, CalCOFl 5106 Sta. 77.65, dorsal view of
8.5 mm specimen at left; (D) B. wesetlu. 1 1.3 mm. from Ahlstrom (1972b), dorsal view of 8.5 mm specimen at left; (E) B. mgrigenys. 21.8 mm,
SIO Shellback Exped. Sta. 92, dorsal view of 8.7 mm specimen at left; (F) B. argyrogaster. 17.1 mm, Dana Sta. 4003.
AHLSTROM ET AL.: ARGENTINOIDEI
165
166
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
AHLSTROM ET AL.: ARGENTINOIDEI
Table 42. Characters used in Analysis of Four Argentinoid Groups.
167
Character
number
Dervied character state
Outgroup
1 Accessory cartilage at posterior tip ceratobr. 5
2 PU, + U, fused
3 Light organs present
4 Frontals fused
5 Epibr. 4 with one post. art. surface
6 Larval gut with stomach
7 Pelvic fins form early and large
8 Swimbladder absent
9 Urodermal absent
10 LL scales extend onto caudal fin
1 1 Larval gut folded
12 Extrascapular attached to pterotic
13 Uncinate process lacking on epibr. 4
14 Pectoral fin forms early and large
Osmerids
Greenwood and Rosen, 1971
Teleosts in general
Goody, 1969
Teleosts in general
Bertelsen and Munk, 1964
Teleosts in general
Cohen, 1964b
Osmerids
Greenwood and Rosen, 1971
Osmerids
This paper, Heame (this volume)
Osmerids
This paper, Heame (this volume)
Teleosts in general
Cohen, 1964b
Teleosts in general
Greenwood and Rosen, 1971
Teleosts in general
Osmerids
Heame (this volume)
Teleosts in general
Chapman, 1942
Osmerids
Greenwood and Rosen, 1971
Osmerids
This paper, Heame (this volume)
of a line of small melanophores above and below the notochord
tip, a patch of melanophores on the opercle and groups of small
melanophores on the upper and lower jaws. Balhylagns anlarcti-
cus has 3 lateral gut spots, a large lateral trunk melanophore at
the I0lh-I2th myomere, and head and notochord pigment sim-
ilar to that of B. euryops. Early larvae of B. pacificus have a
large lateral blotch at mid-body and another one posteriad on
the body. Initially these melanophores are located at the junction
of the gut and body but in later larvae are located on the trunk.
Later a 3rd blotch forms midway between these two. A 4th
lateral trunk blotch forms in some late larval specimens between
the pectoral fin and the large mid-body blotch and melanophores
form lateral to the liver and at the free terminal section of the
gut. Head and notochord pigment is similar to B. euryops and
B. antarcticus.
Bathylagiis hericoides is unusual in having only a series of as
many as 18 lateral gut melanophores (Fig. 86). Late postflexion
larvae develop pigment on the lower jaw, isthmus, opercle, pec-
toral fin base and lateral caudal peduncle. Bathylagus longiros-
tris develops a heavier pattern of pigmentation, beginning with
a series of small melanophores on the posterior section of the
gut in early larvae. Also in preflexion larvae a series of rect-
angular-shaped melanophores develops on the hypaxial myo-
meres. Later in the larval period the lateral gut series is extended
forward along the entire gut, although with wider spacing than
on the posterior gut section. Also, the epaxial myomeres develop
rectangular-shaped melanophores, beginning posteriorly and ac-
cruing anteriorly. The head develops pigmentation from the
opercle to the jaws (Fig. 86). Bathylagus ochotensis develops a
similar pigment pattern except that the melanophores on the
posterior gut section are comparatively larger and fewer, the
anterior region of the gut lacks melanophores and the epaxial
myomere series is limited to the posterior region.
Larvae of B. wesethi. B. nigrigenys and B. argyrogaster have
a similar pigment pattern that differs markedly from that of
other Bathylagus (Fig. 86). Initially there is a series of paired
melanophores dorsolateral to the gut, extending from the pec-
toral fin base to the terminal section. These remain throughout
the larval period but become embedded and obscured in late
larvae. Bathylagus nigrigenys begins with about 8 pairs, which
increase to 10, whereas B. wesethi begins with 6 pairs and has
7-8 during most of the larval period. Both species develop pig-
ment at the notochord tip; B. wesethi has a dorsal and ventral
spot, while B. nigrigenys has only a ventral spot. At notochord
flexion a series of melanophores appears along the hypaxial
region of the body and, soon after, a series develops along the
epaxial myomeres. More lateral series are added and in late
larvae the entire body is covered. Melanophores also form in
the median finfold of advanced larvae. Initially head pigmen-
tation consists of melanophores on the opercle and jaws but in
later larvae the entire head is covered.
Opisthoproctid larvae have distinctive and, in some genera,
heavy pigment patterns (Fig. 87). Bathylychnops has a dorsal
series of 6 large paired blotches that permeate the musculature,
bridge across the longitudinal septum and expand onto the fin-
fold. A series of 8 large ventrolateral blotches alternate with
those of the dorsal series, with the exception that the postanal
blotch lies opposite the dorsal blotch and expands to form a
band. A large blotch covers the base of the caudal fin. The head
IS heavily pigmented with superficial melanophores on the bran-
chiostegals, urohyal and lateral brain and deeply embedded me-
lanophores in the snout, jaws, cheek and ventral brain region.
The lower limbs of the gill arches and their filaments are heavily
pigmented as are both the pectoral and pelvic fin bases.
The species of Dolichopteryx have lateral series of melano-
phores above the gut and some species develop serial melano-
phores on the hypaxial myomeres (Fig. 87). Head pigment con-
sists of melanophores on the jaws, gill arches and, in most species,
the internal snout region. Macropmna develops a series of slant-
ed melanophores, one on each hypaxial myomere, and a heavy
embedded blotch at the pelvic fin base, that expands both dorsad
and ventrad as a band. The caudal fin base has a large blotch
Fig. 87. Larvae of Opisthoproctidae. (A) Bathylychnops exilis. 1 5.6 mm. CalCOFI 7203 Sta. 67.80; (B) Ventral view of above; (C) Dolichopteryx
hinoculans. 58.0 mm, redrawn from Roule and Angel (1930); (D) Afacropinna microstoma. 1 1.7 mm, CalCOFI 7412 Sta. 120.50; (E) Ventral
view of above; (F) Rhynchohyalus natalensis. 23.0 mm, from Bertelsen et al. (1965); (G) Opisthoproctus grimaldii. 14.0 mm from Schmidt (1918).
168
ARGENTINIDAE
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
MICROSTOMATIDAE BATHYLACIDAE OPISTHOPROCTIDAE
3,4,5,6,7,
(13), (14)
Fig. 88. Cladogram showing the distribution of character states in four nominal families of argentinoid fishes. Numbers refer to characters in
Tables 42 and 43. Parentheses mdicate character reversals.
and in the gut region there is pigment above the terminal section
and ventral to the liver. Head pigment is confined to the lower
jaw. The pigment pattern of Rhynchohyalus as described by
Bertelsen et al. (1965) consists of a series of four dusky bars
beginning at the pelvic fin and ending at the caudal fin base.
Embedded beneath these is a layer of diffuse melanophores
which becomes denser toward the caudal. The pectoral fin bases
are pigmented and in the ventral region there are melanophores
on the isthmus and gut. The anal light organ is covered with a
melanistic sheath. The late larval specimen of Opisthoproctus
grimaldii illustrated by Schmidt (1918) shows a diffuse covering
of melanophores over the body and a dusky bar extending down
from the dorsal fin. A 10 mm larva of O. so/eat us in our col-
lection has a pigment pattern similar to Macropuma, with a
series of slanted melanophores on the hypaxial myomeres,
embedded blotches at the pelvic and caudal fin bases, pigment
at the liver and ventrally at the angles of the lower jaw.
Transformation stage
In argentinids transformation from larva to demersal juvenile
is a prolonged process and pelagic juveniles with the retained
larval pigment blotches or bars have been reported many times
(see Cohen, 1958; Nishimura, 1966). Morphological changes
(e.g., deepening of the body, prolongation of the snout, eye
enlargement) and the masking of the larval pigment occur grad-
ually. The beginning of this stage may be defined by the folding
of the anterior gut region to form a stomach. This occurs at 25-
30 mm in Argentina sialis, but has not been documented for
other species. Pelagic juveniles of Glossanodon and A. sialis
develop a silvery stripe at the lateral line region. This has not
been reported for pelagic juveniles of A. silus and A. sphyraena
and may afford an additional character for separating Argentina
into two groups. The end of the pelagic juvenile stage, marked
by the development of scales and silvery integument, is attained
Table 43. Distribution of Char.acter States in Folir Nominal
F.MHILIES OF Argentinoid Fishes. Direction of transformation A * B.
Character
Opistho-
number
Argenlinidae
Micrstomatidae
Bathy[ag]dae
proctidae
1
B
B
B
A
2
B
B
B
A
3
A
A
A
B
4
A
A
A
B
5
A
A
A
B
6
A
A
A
B
7
A
A
A
B
8
A
A
B
A
9
A
A
B
A
10
A
B
A
A
II
A
B
A
A
12
B
A
A
A
13
A
B
B
B
14
A
B
A
B
AHLSTROM ET AL.: ARGENTINOIDEI
169
at various lengths by different species. Schmidt (1906c) reports
complete transformation at about 50 mm in A. sphyraena and
at a much larger size in A. silus. Size at completion of trans-
formation in Glossanodon species is also in the 50-100 mm size
range (Nishimura, 1966).
Microstomatids develop a lustrous guanine layer on the in-
tegument in late larvae and some species develop distinct ju-
venile pigmentation. In Mil rosloma juvenWcs the region of the
body from the dorsal fin origin posteriad is more darkly pig-
mented than the rest of the body, and grades to a solid black
pigment at the caudal fin base. Juveniles of some Nansenia
species develop heavy melanistic pigment at the base of the
caudal fin and often at the base of the adipose fin (Schmidt,
1918; Kawaguchi and Butler, in press).
Bathylagids have a direct transformation and undergo a marked
morphological change from the slender larval form to the ju-
venile form, characterized by a large head and eyes and deeper
body. The gut becomes coiled and covered by a black peritoneal
sheath. The head becomes heavily pigmented but the body is
slower to develop the black pigment characteristic of all Bath-
ylagiis species (other than the B. stilbnis group) and, in species
such as B. euryops and B. nulleri. the large larval melanophores
are visible in specimens up to 30 mm and 50 mm respectively.
In the deep-bodied opisthoproctid genera transformation to
the juvenile stage is marked by deepening of the body and at-
tainment of melanistic integument and large scales. Cohen ( 1 960)
described the large (up to 124 mm) transitional specimens of
Bathylychnops which are semi-transparent and retain the large
larval pigment blotches. Sexually mature specimens of Doli-
chopteryx are semi-transparent, have a membranous body en-
velope, poorly developed musculature, an exposed gut covered
only by peritoneum, weakly attached fins and melanistic pig-
ment of the type usually associated with larvae (Cohen, 1960).
Relationships
Our survey of argentinoid ontogenetic characters provides
insight into some of the systematic questions posed at the be-
ginning of the paper. A close relationship between argentinoids
and alepocephaloids is not supported since the latter hatch from
large eggs (estimated at 3-4 mm based on size of yolk-sac lar-
vae), have direct development, and share no specialized onto-
genetic characters with argentinoids. Four major argentinoid
lineages can be defined by specializations of the eggs and larvae
and thus four families recognized: Argentinidae, Microstoma-
tidae, Bathylagidae, and Opisthoproctidae. Argentina and Glos-
sanodon have generalized larvae except that all known species
have distinct lateral series of melanistic blotches or bands, not
found elsewhere among argentinoids. The pattern of banding
does not separate the two genera.
All known bathylagid eggs have multiple oil globules. A num-
ber of bathylagid groups are apparent from larval characters: 1)
niillcri, 2) slilhms-schmidti-iirotranus, 3) euryops- pad ficus-ant-
arcticus, 4) hericoides-longirostris, 5) wesethi-argyrogaster-ni-
grigenys. Of these groups, stilbius-schmidti-urotranus has the
most generalized morphology and pigmentation, lending no
support for its separation as a distinct genus.
Opisthoproctid larvae share a number of neotenic features,
including a saccular stomach. Except for body shape, Dolichop-
teryx shares more derived larval characters with the deep-bodied
genera than with Bathylychnops. and the latter has a number of
characters unique to opisthoproctids. Division of the family
based on body shape is not supported by ontogenetic evidence.
Ontogeny offers little information on species composition of
genera, because only a fraction of argentinoid eggs and larvae
are known. However, egg and larval characters clearly separate
Atlantic and Pacific Microstoma as distinct species. Bathylagits
hericoides larvae from the Atlantic and Pacific are indistinguish-
able. The same is true for B. longirostris from all oceans. Bath-
ylagus nigrigenys and B. argyrogaster larvae are indistinguish-
able, lending support for a single circumtropical species.
Bathylagus stilbiiis eggs and larvae are indistinguishable from
those of B. urotranus.
We have attempted to analyze the distribution among four
nominal groups of argentinoids, of 14 characters, four of which
are taken from developmental stages and 10 from the adult
(Table 42). We have used teleosts in general and osmerids as
our outgroup following Fink and Weitzman ( 1 982). Distribution
of character states are presented in Table 43.
A possible arrangement of groups based on the fewest number
of character reversals is presented in Figure 88. Opisthoproc-
tidae appears to be a well-founded family. More precise inter-
pretation of the inter-relationships and nomenclatural ranking
for argentinids, microstomatids, and bathylagids requires ad-
ditional data.
(H.G.M.) Southwest Fisheries Center, P.O. Box 271, La
JoLLA, California 92038; (D.M.C.) Natural History
Museum Los Angeles County 900 E.xposition Boule-
vard, Los Angeles, California 90036.
Stomiatoidea: Development
K. Kawaguchi and H. G. Moser
FISHES of this group of midwater predators are characterized
by their dark coloration, serial photophores, large jaws,
fang-like teeth, and chin barbels. Traditionally they have been
grouped in six families allied to the lightfishes and hatchetfishes
(Weitzman, 1974), and together are now considered monophy-
Ictic and given ordinal status (Rosen, 1 973; Fink and Weitzman,
1982). Fink (this volume) gives evidence for reducing the six
stomiatoid families to one. Because knowledge of stomiatoid
ontogeny lags far behind that of the adults, for convenience of
discussion we use Weitzman's (1974) grouping of the families
170
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 89. Larvae of Slomias and Chauliodus. (A) 5. alnventer. 4.6 mm, CalCOFI 7501 Sta. 97.60; (B) 5. atriventer, 10.0 mm, CalCOH 6604
Sta. 107.65; (C) 5. atmenler. 22.2 mm; CalCOH 6604 Sta. 107.65; (D) S. ferox. 30 mm, from Ege, (1918); (E) C. sham. 6.0 mm; from Mito
(1961a); (F) C macouni. 15.0 mm, CalCOH 6204 Sta. 60.60; (G) C. macouni. 45.2 mm, CalCOH 5707 Sta. 67.60.
Astronesthidae, Stomiatidae, Chauliodontidae. Melanostomia-
tidae, Maiacosteidae, and Idiacanthidae, in the Superfamily Sto-
miatoidea.
Eggs
Eggs are known for Chauliodus, Stomias, and Tactostoma
and have in common a round shape, smooth chorion, and seg-
mented yolk, Chauliodus eggs have a wide perivitelline space
and lack an oil globule. Egg diameters are: C sloani. 2.2-2.5
mm (Sanzo, 1931d); C. barbatus, 3.1-3.6 mm (Pertseva-Os-
troumova and Rass, 1973); C macouni. 2.7-3.1 mm, with an
initial yolk diameter of 1.3-1.5 mm (original data). Mito (196 la)
described an egg, referred to C. sloani. l.\l mm in diameter
with no oil globule but with a second membrane. Stomias eggs
have a second membrane, a single oil globule and the following
diameters: S. colubrinus. 1.3-1.5 mm, with inner membrane
1.05-1.1 mm (Pertseva-Ostroumova and Rass, 1973); S. atri-
venter. 0.88-0.92 mm, inner membrane diameter is 0.82-0.84
KAWAGUCHI AND MOSER: STOMIATOIDEA
171
Table 44. Meristic Counts of Stomiatoid Genera. Most frequent count or range is followed by overall range or infrequent count in parentheses.
Data from Gibbs (1964a,b), Gibbs et al. (1983), Morrow (1964a, b, c). Morrow and Gibbs (1964), Bolin (1939a), Imai (1941). onginal counts.
Vertebrae
Fin rays
Family and genus
Dorsal
Anal
Pectoral
Pelvic
Stomiatidae
Macrostomias
164
13,14
16(15-18)
7(6)
4
Stomias
64-83
17-20(16-22)
19-21 (18-25)
6-7 (6-9)
5(4)
Chauliodontidae
Chautiodus
51-62
6,7 (5-7)
10-12(10-13)
12,13(11-14)
7 (6-8)
Astronesthidae
Astronesthes
46-58
15(10-21)
12-22
8 (5-9)
7 (6-8)
Borostomias
53-55
13(10-14)
13-16(10-19)
7 (6-9)
7
Hetempholus
66
11 (13)
12-15(17)
7
7
Neoneslhes
53
9-11 (12)
25-27 (22-28)
8(7)
7 (6-8)
Rhadineslhes
67
11 (12.13)
18(19-21)
7 (6-8)
7
Melanostomiatidae
Bathophilus
38-45 (33-50)
13-16(9-18)
15-16(9-18)
1-37
11-16(4-26)
Chiroslomias
54-55
18-20
22-26
6
7
Echwstoma
57-59
11-14(11-16)
13-18(13-19)
1 + 3
8
Eiistomias
56-69
21-25(20-30)
32-46
0-13
7 (6-8)
Ftagellostomias
65
16(14-17)
23-25(21-26)
I + 8-9 + I1
7
Grainmalostomias
50-56
18-21
21-23(20-24)
4-11
7-8
Leptostomias
77-80 (75-83)
16-22
20-29
10(9-11)
7(8)
Melanostomias
50-57
12-17
16-20
5(4-6)
7(8)
Opostoimas
60
21
24
1+4
8
Pachystomias
48
22(21-24)
27 (25-29)
5-6
8-9 (7)
Photonectes
49-64
15-24
17-24
0-3
7(6)
Tactosloma
80-82
14-16
19-22
0
8-10
Thysanaclis
61
17-18
21-25
1+10,11
7
Trigonolampa
61-62
19-20(18)
18(19)
5
7
Malacosteidae
Anslostomias
44-56
18-26
24-32
6-10(3-17)
6
Malacosteus
49
14-19(20)
17-21 (23)
3-4 (5)
6
Photostomias
52-58
22-28
25-32
0
6
Idiacanthidae
Idiacamhus
79-85
54-74
34 (33-39)
0
6
mm, oil globule diameter of 0.20-0.25 mm, initial yolk diameter
of 0.70 mm (original data). Tactosloma macropus eggs have a
single membrane, 1.44-1.54 mm in diameter, an oil globule
0.30-0.40 mm in diameter and an initial yolk diameter of 0.78-
0.80 mm (original data). Eggs of C. macouni and 5. athventer
are illustrated in Matarese and Sandknop (this volume).
Larvae
Larvae of Stomiatoidea occur in the upper water column,
some at the surface. In most groups the larvae are elongate, have
a large head, elliptical eyes that protrude slightly from the dorsal
head profile, an elongate, straight gut (trailing from the body in
some species), a well developed finfold, large paddle-shaped
pectoral fins that lack rays until transformation, and late-form-
ing pelvic fins. Melanophore patterns provide a useful set of
characters and genera usually have a distinct pattern. The larval
melanophores are retained in a subcutaneous position in trans-
forming specimens and provide a means for identifying larvae.
During transformation, photophores form simultaneously and
initially are unpigmented. Counts of fin rays, vertebrae, and
photophores are summarized in Tables 44 and 45.
Stomiatidae (Fig. 89). — Larvae of five species are known (Table
46). Larvae are 3-4 mm at hatching and have an elongate yolk
sac. The slender body is round in cross-section, but becomes
slightly deeper by late postflexion. The head is relatively small
with a slightly flattened snout. The eyes are elliptical. The elon-
gate gut extends almost the entire length of the body and has a
slightly enlarged terminal section that reaches the anal fin origin.
The median finfold is small and best developed posteriorly. The
opposing dorsal and anal fins develop far posteriad on the body
in early postflexion larvae, but the pelvic fins do not appear
until just before transformation.
Late-stage embryos oi Stomias have melanophores along the
dorsum, which migrate ventrad and form a distinct series be-
tween the body and gut. This series extends to the tip of the
notochord. The series is lost before notochord flexion but, in
most species, another sparser series develops along the ventral
midline of the gut, from the isthmus to the anus. 5". boa and S.
fero.x develop a mid-lateral series of melanophores along the
body and S. colubrimis has scattered melanophores along the
entire hypaxial body region. These species also develop exten-
sive dorsal and lateral head pigment. All species form scattered
pigment on the dorsal, anal, and caudal fins.
A 75-mm specimen (MCZ Cat. No. 59858) with an extremely
slender body form (body depth 1.3% of body length) has fin and
172
ONTOGEIVY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
Table 45. Photophore Counts of Stomiatoid Genera. Most frequent count or range is followed by overall range or infrequent count in
parentheses. Data sources as in Table 1. Photophore groups as defined by Morrow (1964a).
Photophore
groups
Family and genus
IP
PV
vav
AC
ov
VAL
Stomiatidae
Macrostomias
11(12)
80-86
58-67
19-22
79-85
58-68
Stomias
9-13
32-51
5-16
14-20
32-50
4-17
Chauliodontidae
Chauliodus
8-11
17-23
22-30
8-13
17-21
22-29
Astronesthidae
Astronesthes
5-12
6-20
7-27
7-13
5-19
7-26
Borostomias
10-13
20-31
15-25
9-15
21-29
16-25
Heterophotus
10-11
32-35
13-14
12-15
33-36
16-20
Neonesthes
9-12
14-17
16-21
13-18
13-15
13-21
Rhadinesthes
10(6)
25(26)
20-23
16
22-24
27
Melanostomiatidae
Bathophiius
5 (4-6)
12-18
11-13(11-
17)
5-7 (5-9)
13-14(10-16)
9-11 (8-17)
Chirostomias
9(8)
25-27 (28)
19-20(16)
9(10)
23 (24-25)
19-20(16)
Echiostoma
8 + 2
25-28
14-18
12-13(11)
24-31
13-17(18)
Eustomias
7-8 (9)
27-33 (24-36)
13-17(11-
21)
17-23(15-25)
26-33 (24-37)
13-18(12-22)
Flagellostomias
9-10(8)
31-34
14-16
16-18(15)
31-32(30)
14-15(12-17)
Grammatostomias
7(6)
15-18
19-22
10-13
15-18
19-22
Leptostomias
10(11)
42-45 (39-48)
20-23 (24)
11-13(14)
40-43 (39-48)
20-22 (23-24)
Metanostomias
8 + 2 or 3
23-30
12-15
9-11
22-28
11-15
Opostomias
4 + 4
27
17
16
27
17
Pachysiomias
8-9
14-16(17)
13-14
8-9
17-18
14-15
Photonecies
8-11
19-24, 34-38
11-15(16-
18)
10-13(9)
19-24(17), 30-36
11-14 (15-17)
Tactostoma
8
46
19
12
43
18
Thysanactis
20
31-32
14-16
11-12
30-32
14-16
Trigonolampa
11
23-24 (22)
22 (24)
10-11
22-24
23-24 (26)
Malacosteidae
Aristoslomias
5 + 3
15-17(14-19)
15-18
9-11 (12)
16-19(14-20)
15-17(14-18)
Malacosieus
(Serial photophores absent or uncountable)
Photostomias
5 + 2
13-16
21-25
12-15
12-17
20-23
Idiacanthidae
Idiacanlhus
1P + PV = 31-36
16-18(15)
13-18
22-25
31-35(30-36)
vertebral counts that match Macrostomias longibarbatus. Its
morphology is that of a highly attenuate Stomias larva. Pig-
mentation is restiicted to a series of small melanophores along
the ventral midline of the gut. The ventral photophore rows are
beginning to form.
Chauliodontidae (Fig. 59). — Larvae of five species are known
(Table 46). Larvae are 6-7 mm long at hatching, with an elongate
yolk sac. The body is slender with a circular cross-section, and
remains so throughout development. The head is relatively small,
with elliptical eyes and a short, acute snout. The gut has a smaller
diameter than in Stomias but is relatively longer. The short
terminal section extends beyond the anal fin origin. The median
finfold is small and best developed rearward on the body. The
dorsal, anal, and pelvic fins form in late postflexion larvae in
the adult position. A fan-shaped array of melanophores occurs
in the caudal region of yolk-sac larvae but is soon lost. No other
pigment develops. Larvae of some species reach 46 mm SL and
there appears to be marked shrinkage at transformation.
Astronesthidae (Fig. 90). — Astronesthid larvae have been illus-
trated and described briefly by Roule and Angel ( 1930), Whitley
(1941), Pertseva-Ostroumova and Rass (1973), and Belyanina
(1982b); however only two of these were identified to genus
(Table 46). We have examined more than 10 types of astro-
nesthid larvae, 7 of which are listed in Table 46. Astronesthid
larvae display a great variety of structure and pigmentation, but
hold in common the advanced position of the dorsal fin, in
contrast to other Stomiatoidea, except Chauliodus. The types
differ fundamentally in gut shape and body form: Types I and
II are laterally compressed, relatively deep-bodied, and have a
non-trailing or slightly trailing gut with terminal section as in
melanostomialids; Types III-VIl have a slender body and a
trailing gut; in Types III-V the gut is deflected ventrad from
the body just anterior to the anal fin base and in Type VI and
VII at midbody, anterior to the dorsal fin (Figure 90).
Type I (Fig. 90A). — larvae up to 26.5 mm; laterally compressed;
head shallow with acute snout; eyes relatively large, slightly
Fig. 90. Larvae of Astronesthidae. (A) Type I, 23.7 mm, ORl A105; (B) Type II. SIO Tasaday I A3; (C) Type IV. 33.0 mm. MCZ Cat. No.
59855; (D) Type V, 22.0 mm, Dana Sta. 3931; (E) Type VII, 28 mm, MCZ Cat. No. 59856.
KAWAGUCHI AND MOSER: STOMIATOIDEA
173
174
ONTOGE^Pr' AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 46. Pigment Characters and Gut Structure in Larvae and Transforming Specimens of Stomiatoidea. (NT = not trailing, ST =
slightly trailing, T = trailing freely).
Hypaxial
myoseptum
Length
Length of
Dorsal myomere
Epaxial myoseptum
melanophores
of larvae
ti^nsfonning
melanophores
melanophores
(no./myo-
Gut
Species
(mm)
specimens (mm)
(no./myomere)
(no. /myoseptum)
seplum)
stnicture
Source
Stomiatidae
Stomias boa
_
38
0
0
0
NT
Sanzo, 1912a
Stomias boa
10.4-30.4
41.5
0
0
0
NT
Sanzo, 193 Id
Stomias boa
9.0-32
—
0
0
0
NT
Ege, 1918
Stomias ferox
9.0-44
—
0
0
0
NT
Ege, 1918
Stomias colubrinus
3.3-16
—
0
0
0
NT
Pertseva-Ostroumova
and Rass, 1973
Stomias alriventer
4.6-32
—
0
0
0
NT
original
Macrostomias longibarbatus
-
75
0
0
0
NT
original
Chauliodontidae
Chauliodus sloani
33.6
41.6
0
0
0
NT
Sanzo, 1915a
Chauliodus sloani
5.7-41.6
27.1
0
0
0
NT
Sanzo, 193 Id
Chauliodus sloani
2.1
—
0
0
0
NT
Mito, 1961a
Chauliodus danae
22.5-25
—
0
0
0
NT
Belyanina. 1977
Chauliodus macouni
38.0-49
35-44
0
0
0
NT
Belyanina, 1977
Chauliodus mmimus
23.5-35
—
0
0
0
NT
Belyanina, 1977
Chauliodus pammelas
10.6-40
_
0
0
0
NT
Belyanina, 1977
Chauliodus sloani
7.4-35
27-34.2
0
0
0
NT
Belyanina, 1977
Chauliodus macouni
5.6-46
-
0
0
0
NT
original
Astronesthidae
Unidentified
14.0-23
—
0
0
0
T, NT
Roule and Angel, 1930
Astronesthes lupina
20
—
0
0
0
T
Whitley, 1941
Boroslomias panamense
5.0-17
—
0
0
0
T
Pertseva-Ostroumova
and Rass, 1973
Unidentified
16
—
7
+
+
NT
Belyanina, 1982b
Unidentified
17.7
—
2 total
0
0
T
Belyanina, 1982b
Type I
12.3-26.5
—
several
several
to many
several
to many
NT
original
Type II
14.9-26
29,40
0
0
0
ST
original
Type III
16.2
20.5,22.5
0
0
0
T
original
Type IV
14.4-34.5
40.5
0
0
0
T
original
TypeV
17.4-19.4
20,22
0
0
0
T
original
Type VI
—
28
0
0
0
T
original
Type VII
—
28
0
0
0
T
original
Melanostomiatidae
Tactostoma macropus
5.0-44
49
0-1
0
1-3
NT
original
Melanostomias spilorhynchus
17
21-32
0
ca. 3
NT
Beebeand Crane, 1939
Melanostomias biseriatus
—
21-25
0
ca. 3
NT
Beebe and Crane, 1939
Melanostomias valdiviae
—
25
0
2-3
NT
original
Melanostomias sp.
13.4-17.2
16.4-22
0
2-4
NT
original
Echiosloma tanneri
20,25
—
0
2-5
NT
Beebe and Crane, 1939
Echiostoma sp.?
13.8
_
0
2-4
NT
Belyanina, 1982b
Echiostoma barbatum
—
34
0
1-2
NT
original
Photonectes dinema
—
24 and >
1 (?)
0
3-4
NT
Beebe and Crane, 1939
Photonectes leucospilus
—
25 and >
1(?)
0
3-4
NT
Beebe and Crane, 1939
Photonectes albipinnis
—
16-22
0
2-3
NT
original
Photonectes sp.
11.0-12.5
—
0
4-5
NT
original
Photonectes parvimanus
12.0-26
25
3-6
0
3-4
NT
Beebe and Crane, 1 939
Photonectes parvimanus
27
—
3-4
0
2-4
NT
original
Photonectes parvimanus
—
28
1-2
0
2-4
NT
original
Photonectes sp.
5.4-22.2
—
ca. 7
0
5-7
NT
original
Opostomias mitsuii
15.0-21
1
0-1 (2-3
posteri-
orly)
1-2 (3-5
post.)
NT
original
Flagellostomias boureei
20.0-21
34,39
1
0
1
NT
Beebeand Crane, 1939
Flagellostomias boureei
10.8-36.4
—
1
0
1-2
NT
original
Odontostomias micropogon
—
42
1
1-?
2-4
NT
Beebe and Crane, 1939
Leptostomias gladiator
12.0-30
38-45
1 + several
1-5
2-4
NT
Beebe and Crane, 1939
Lepiostomias gracilis
—
37.8
1 + 1-5
5-7
6-9
NT
original
Leptostomias sp.
25
—
1 + 1-3
4-5
4-6
NT
original
Bathophilus nigerrimus
11.6
21.7
1 or >
0
0
NT
Sanzo, 1915a
Bathophilus nigerrimus
5.9, 14.0
19.2-21.7
1 or >
0
0
NT
Sanzo, 193 Id
KAWAGUCHI AND MOSER: STOMIATOIDEA
175
Table 46. Continued.
Hypaxial
myoseptum
Length
Length of
DoPial myomere
Epaxial myoseptum
melanophores
of larvae
transforming
melanophores
melanophores
(no./myo-
Gut
Species
(mm)
specimens (mm)
(no, myomere)
(no./myoseptum)
septum)
structure
Source
Bathophilus metallicus
25
3 or >
0
0
NT
Beebe and Crane, 1939
Bathophilus sp.
11, 12
1 or >
0
0
NT
Beebe and Crane, 1939
Bathophilus sp.
7
1 or >
0
0
NT
Beebe and Crane, 1939
Bathophilus sp.
15
_
(?)
0
0
NT
Rouleand Angel. 1930
Bathophilus sp.
18.2
—
1 or >
0
0
NT
de Sylva and Scotten,
1972
Bathophilus filifer
4-10
-
1 or >
0
0
NT
Pertseva-Ostroumova
and Rass 1973
Bathophilus brevis
15.7
—
1 or >
0
0
NT
original
Bathophilus Jlemingi
2.9-23.8
—
1 to several
0
0
NT
original
Euslomias sp.
33
_
7 total
0
0
T
Regan, 1916
Eustomias sp.
13
—
7 total
0
0
T
Beebe and Crane, 1939
Eustomias spp. (4 types)
6.0-45
-
5-1 1 total
0
0
T
original
Malacosteidae
Aristostomias scintillans
4.3-47
45
14 total
to many
0
0
T
original
Photostomias guernei
20.0-27.5
30,31
8 pairs
total
0
0
T
original
Unidentified
12
_
1 2 total
0
0
T
Beebe and Crane, 1939
Unidentified
34.5
-
0
0
0
T
original
Idiacanthidae
Idiacanthus fasciola
16.0-28
35-48
0
0
1
T
Beebe, 1934
Idiacanthus sp.
7.0-39
—
0
0
1
T
Pertseva-Ostroumova
and Rass, 1973
Idiacanthus antrostomus
4.5-71
67->
0
0
1
T
original
elliptical; gut moderately slender, thin-walled; finfold moderate;
pigment pattern consists entirely of minute melanophores, in-
creasing in number with development, principally in the ex-
paxial and hypaxial myosepta; other pigment above brain, paired
internal streaks in snout, melanophores in dorsal and ventral
finfold, dorsal fin base, and on posterior half of gut.
Type II (Fig. 90B). — larvae reach at least 26 mm; deep-bodied
and laterally compressed in late-stage larvae; head deep; eyes
small, slightly elliptical; gut slightly trailing and with larger di-
ameter than in Type I; dorsal finfold relatively deep; pigment
above brain, along lower jaw and at angular and gular region;
blotch at posterior margin of superior hypural complex and one
midway out on inferior group of caudal rays; fin ray and ver-
tebral counts and photophore counts match Astronesthes gem-
mifer.
Type III. — larvae reach at least 16.2 mm; body slender; head
and eyes moderate in size; eyes elliptical; slender gut trails free
from body at anal fin origin; finfold moderately developed, ex-
cept posterior to dorsal fin the finfold appears as an enlarged
adipose fin; pigment restricted to a series of melanophores along
lower jaw and between upper and lower hypural complexes;
counts match Astronesthes richardsoni.
Type IV (Fig. 90C).— lai^ae reach 40 mm; morphology similar
to Type III, except head relatively longer and eyes almost round;
gut with leaf-like appendages on trailing section; pigment re-
stricted to postorbital blotch and interorbital band; fin and ver-
tebral counts and photophore arrangement match Heterophotus.
Type V (Fig. 90D). — larvae reach about 20 mm; morphology
as in Types III and IV; eyes slightly elliptical; pigment heavy;
melanophores on head, lateral to posterior brain region, on
snout and lower jaw symphysis; lateral surface of body covered
with an irregular pattern of large melanophores; melanophores
on trailing gut. Pertseva-Ostroumova and Rass (1973) identified
larvae of this type as Borostomias panamense.
Type VI. — specimen transfoiming at 28 mm; morphology sim-
ilar to Types II-V, except trailing gut deflected from body far
in advance of anal fin origin; eyes elliptical; dorsal finfold highly
developed and ventral finfold anterior to anal fin is rudder-like;
pigment lacking; meristics indicate it is in the genus Astro-
nesthes.
Type VII (Fig. 90E). — specimen transforming at 28 mm; mor-
phology similar to Type VI; dorsal and anal fins supported on
cartilaginous pedestals; a series of 4 melanophores along hori-
zontal septum; some melanophores on anterior region of dorsal
and anal fin bases and on preanal finfold. Whitley (1941) de-
scribed a larva similar to this as Astronesthes lupina.
Melanostomiatidae (Figs. 91-92). — Larvae have been identified
for 10 of the 15 genera (Table 46). Bathophilus was the first to
be identified (Sanzo, 1915a). The only comprehensive work on
melanostomiatid ontogeny is that of Beebe and Crane (1939)
who identified larvae of 8 genera and 5 species by the use of
transforming series. Since then, the only other melanostomiatid
larvae that have been described are Bathophilus filifer {Pertseva-
Ostroumova and Rass, 1973), Bathophilus sp. (de Sylva and
Scotten, 1972), and Echiosloma (?) sp. (Belyanina, 1982b). De-
176
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
scriptions of Opostomias and Tactostoma are included in this
paper. Larvae of Tactostoma were initially identified by E. H.
Ahlstrom.
Larval representatives of the 10 genera are highly various in
form and pigmentation, however, with the exception of Euslo-
mias, they share the following structural features: body elliptical
in cross-section; head laterally compressed; eyes small and el-
liptical; gut terminated in an elongate muscular bulb that may
extend beyond the anal fin origin but not beyond the margin of
the finfold; dorsal and anal fins form in adult position posteriorly
on the body; body pigment consists of one or more melano-
phores dorsal to each myomere, one or more melanophores on
the hypaxial myosepta and, in some genera, on the epaxial my-
osepta. Dorsal and lateral pigmentation tends to be heavier in
forms with higher meristic counts. The genera differ principally
in body size, relative body depth, relative head size, jaw size,
gut diameter, size and shape of the terminal gut section, finfold
height, and pigment pattern.
Present knowledge indicates that genera apparently have dis-
tinct facies, tentative descriptions of which are presented below.
Confirmation awaits identification of additional species.
Tactostoma (Fig. 91 A). — larvae reach 44 mm in length; body
extremely slender; head flat and elongate initially, becoming less
flat and relatively smaller with development; eye size moderate;
gut slender; finfold moderate; pectoral fin lost at transformation;
early larvae develop one melanophore per myomere along dor-
sum and 1-3 melanophores on the hypaxial myosepta; post-
flexion larvae gradually lose the dorsal melanophores and then
the hypaxial myosepta pigment, in contrast with other genera
in which body pigment increases with development; pigment
on lower jaw symphysis, isthmus, pectoral fin base, cleithrum,
and above gut terminus; dorsal and ventral pigment accentuated
at caudal peduncle.
Melanostomias (Fig. 91B).— transforming specimens as small
as 16.4 mm; body slender; head small; snout short; eye size
moderate; gut slender; finfold relatively small; one melanophore
per myomere along dorsum in one form and in another form
the zone between the 7th- 10th myomere and the dorsal fin lacks
dorsal pigment; 2-3 melanophores in hypaxial myosepta; pig-
ment above and below head, below liver, on terminal gut sec-
tion, and along finfold margins. Larvae tentatively identified as
Echiostoma have similar characters (Table 46).
Photonectes (Fig. 9 IC). — larvae of different forms transform at
sizes between 16 and 28 mm; body somewhat deep; head size
and snout length moderate; eyes small, highly elliptical; several
forms of dorsal myomere pigment ( 1 melanophore per myomere
in Subgenus Photonectes and 3-7 per myomere in Subgenus
Trachinostomias); hypaxial myosepta with 2-7 melanophores
depending on form (Table 46); extensive pattern of minute me-
lanophores on head, finfold, and median fins.
Flagellostomias (¥\g. 9 ID). — larvae may reach 30-40 mm; body
somewhat deep; head large, deep, with steeply sloping snout and
large jaws; eyes small; gut diameter relatively large; finfolds
large, accentuating body depth; one large melanophore per myo-
mere along dorsum; 1-3 melanophores in hypaxial myosepta;
some scattered lateral melanophores in median fin region; other
pigment scant; a few melanophores in head region, some on
finfold in posterior gut region, and on dorsal and anal fins.
Opostomias (Fig. 9 IE). — body moderately deep; head large, deep
posteriorly with elongate sloping snout; eyes small; gut slender;
finfold large; one melanophore per myomere along dorsum; 1-
2 melanophores in hypaxial myosepta; epaxial and hypaxial
myosepta below dorsal fin base have several melanophores,
giving this region a banded appearance; melanophores on dorsal
head region, gill arch and gut terminus.
Leplostomias (Fig. 91F). — larvae may reach about 40 mm; body
somewhat deep; head moderately large, deep; eyes small; gut
slender; finfold moderate; pigmentation heavy; one large me-
lanophore and 1-5 smaller ones per myomere along dorsum;
numerous melanophores on epaxial and hypaxial myosepta,
increasing with development to completely outline myosepta;
pigment extensive on dorsal and ventral head regions, on gill
arches; pigment below liver, on finfold margins, above gut ter-
minus and on dorsal and anal fins.
Bathophilus (Figs. 92A-C). — larvae transform at 25 mm or less;
deep-bodied compared with other genera; head and jaws large;
barbel forms in late postflexion larvae, particularly in B. hrevis;
eye size moderate; gut large to voluminous, with highly devel-
oped s-shaped terminal section; finfolds, particularly dorsal, large;
one or several melanophores per myomere along dorsum and
an opposing series of melanophores along ventral surface of
myomeres; no lateral pigment; head, finfolds and median fins
pigmented.
Eustomias (Fig. 92D). — larvae of some species reach 45 mm;
body slender, and round in cross-section; head elongate and flat
with large spatulate snout; large jaws; eyes moderate in size,
slightly elliptical to round; gut slender, deflected ventrad at anal
fin origin and trailing from body; body pigment consists of 5-
1 1 large melanophores along the dorsal midline; usually pigment
at lower jaw symphysis.
Malacosteidae (Fig. 9iA — Larvae of this group have not been
described, although the 12-mm larva illustrated by Beebe and
Crane (1939) and referred to "lEustomias" is apparently Ar-
istostomias. We have examined larval series and transforming
specimens of A. scintillans and Photostomias guernei (Table 46).
Aristostomias scintillans (Fig. 93A). —larvae reach 47 mm length;
body slender; head large, flat; snout elongate; jaws large; eyes
slightly elliptical; opercle markedly reduced; gut slender, de-
flected ventrad at anal fin origin and trailing from body; finfold
moderate; dorsal and anal fins form in adult position at about
flexion stage; pelvics form late; initial pigment pattern is a series
of paired melanophores along the dorsum, beginning with 14
Fig. 91. Larvae of Melanostomiatidae. (A) Tactostoma macropiis, CalCOFI Norpac Sta. 14; (B) Melanostomias sp., 16.0 mm, ORI KH73-
2, Sta. 49-7; (C) Photonectes sp., 22.2 mm, SWFC, Albacore Oceanography Cruise 71, Sta. 99; (D) Ftagetloslomias boureii. 36.4 mm. SIO Cat.
No. 73-329, Tasaday I, Tow 42; (E) Opostomias mitsiiii. 1 5.0 mm, ORI KH 73-2 Sta. 2-3; (F) Leptoslomias sp., 24.5 mm, MCZ Cat No. 59857.
KAWAGUCHI AND MOSER: STOMIATOIDEA
177
178
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
KAWAGUCHI AND MOSER: STOMIATOIDEA
179
Fig. 93. Larvae of Malacosteidae. (A) Aristostomias scinlillans. 34.7 mm. CalCOFI 5008 Sta. 70.30; (B) Photostomias sp., 26.7 mm. ORI KH
73-5 Sla. 55-13. Bn 24-12; (C) Malacosteidae, 34.5 mm. from Moser (1981).
Fig. 92. Larvae of Melanostomiatidae. (A) Bathophilus flemingi. 25.5 mm. CalCOFI 4910. Sta. 80.137; (B) B hrevis. 15.7 mm. ORI KH
81-1, Sta. 17; (C) B. nigernmus, 21.7 mm, redrawn from Sanzo (1931d); (D) Eustomias sp. 33 mm, redrawn from Regan (1916).
180
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 94. Larva of Idiacanlhus anirostoinus. 55 mm. CalCOFI 6207 Sta. 90.120.
KAWAGUCHI AND MOSER: STOMIATOIDEA
181
pairs and increasing in numbers with development to cover the
entire dorsum; paired ventral series develop, initially poste-
riorly, and increase in numbers so that all myomeres have me-
lanophores on the ventral surface; pigment on brain, snout,
lower jaw, gular-isthmus region, otic region, caudal fin, and in
vague rings along trailing gut. Ahstostomias larvae were iden-
tified initially by E. H. Ahlstrom.
Photostomias giternei (Fig. 93B). — larvae reach about 30 mm;
morphology similar to A. scintillans except eyes smaller and
narrower and pelvic fins somewhat elongate; body pigment con-
sists of a series of 8 minute dorsal melanophore pairs and 8
slightly larger opposing pairs along the ventral surfaces of the
myomeres; melanophores at lower jaw symphysis, large mela-
nophore on each pectoral fin base, sparse melanistic rings along
trailing gut.
Malacosteid C (Fig. 93C). — intact specimen (captured by Dr.
Richard Harbison, WHOl) has morphological and meristic
characters of malacosteid larvae but lacks pigment except on
the extensive gut. Shallow capture locality of this specimen and
our capture of large A. scintillans larvae in MANTA nets in-
dicates late-stage malacosteid larvae have a shallow distribution
in the water column.
Idiacanihidae (Fig. 94j. — Brauer ( 1 906, 1 908) described the re-
markable larvae of Idiacanthus and named them Slylophthal-
mus paradoxus. Beebe ( 1 934) correctly identified the larvae and
described them in detail. Idiacanthus larvae are extremely slen-
der, reaching a length of 35-70 mm depending on the species.
Other characteristics are: elongate and extremely flat head; el-
liptical eyes on long stalks with cartilaginous supporting rods;
stalk length up to 27% of body length in /. antrostonms (Weihs
and Moser, 1981); gut slender, deflected at anal fin origin and
trailing; finfold small; dorsal fin begins forming in preflexion
larvae; dorsal fin larger than anal fin and slightly in advance of
it in postfiexion larvae; during transformation, rays added se-
quentially anteriad so that in adults the dorsal extends about -A
of the body length and the anal about 'A; pectoral fins well
developed but lost at transformation and pelvic fins develop in
transforming females, but not at all in males; pigment pattern
consists of a melanophore on the posterior margin of each hy-
paxial myomere, spreading into the myosepta when expanded,
several elongate internal blotches in the isthmus region, and a
series of melanophores along the trailing gut; adult males of /.
fasciola reach 32-42 mm SL, lack teeth and paired fins and have
relatively larger eyes and an enormous luminous gland.
Relationships
Information on larval characters of 18 of the 26 stomiatoid
genera recognized by Fink (this volume), representing all 6 of
the families recognized by Weitzman (1974), permits some pre-
liminary generalizations and conclusions: (1) Larvae of Sto-
miatidae and Chauliodontidae are similar in morphology and
are distinct from other stomiatoids. Pigmentation provides fur-
ther evidence of this; Chaidiodus larvae are unique among known
stomiatoids in lacking pigment after the yolk-sac stage and the
median series of gut melanophores of Stomias also appear to
be unique. (2) Larvae of Astronesthidae are diverse in mor-
phology and pigmentation and most of the larval specializations
that appear in other stomiatoid families are found among as-
tronesthid genera. Larval specializations of some genera (e.g.,
ornamented trailing gut, trailing gut deflected at mid-body, rud-
der-like finfolds) are not found elsewhere in Stomiatoidea. Het-
erogeneity of larval characters in Astronesthidae supports Fink's
view that the group is paraphyletic. (3) In the Melanostomia-
tidae, larvae of Melanostomias. Photonectes. Echiostoma.
Oposlomias. Flagellostomias. Odontostomias and Leptostomias
are similar in morphology, have paired melanophore series on
the dorsum, and differ chiefly in head size, body depth, and in
the extent of myosepta pigment. Tactostoma larvae have the
characters of this group of genera except that the body is ex-
tremely slender and the pigmentation is lost in the postfiexion
stage. Larvae of Battiophilus difler from those of the above group
in a number of characters (voluminous gut with specialized
terminal section, melanophore series on the ventral surface of
the myomeres, lack of myosepta pigment). Larvae of Eustomias
are different from all known larvae of Melanostomiatidae in
having a trailing gut, flat head and snout, and a pigment pattern
consisting of a median series of up to 11 large melanophores
on the dorsum. Except for this latter feature, Eustomias larvae
are similar to those of Malacosteidae. (4) Idiacanthus larvae
have a combination of characters unique among stomiatoids.
The stalked eyes are autapomorphic. Larval characters provide
no support for Fink's hypothesis that this genus is closely related
to Tactostoma.
Ocean Research Institute, University of Tokyo, 1-15-1,
MiNAMiDAi, Nakano-ku, Tokyo 164, Japan, and Na-
tional Marine Fisheries Service, Southwest Fisheries
Center, 8604 La Jolla Shores Drive, La Jolla, Calif-
ornia 92038.
Stomii forms: Relationships
W. L. Fink
STOMIIFORMS are well known as a major component of
the midwater oceanic fauna. Past concepts of their rela-
tionships to other primitive euleleosts were reviewed by Fink
and Weitzman (1982), but in brief in this century, they have
been considered isospondyls (Parr, 1927; Regan, 1923; Morrow,
1964) or, more recently, salmoniform protacanthopterygians
(Greenwood et al., 1966). In 1973, Rosen placed these fishes as
a separate order (Stomiatiformes) within the Neoteleostei, as
182
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
STERNOPTYCHIDAE
GONOSTOMATIDAE
PHOTICHTHYIDAE
ASTRONESTHIDAE
IDIACANTHIDAE
MALACOSTEIDAE
MELANOSTOMIIDAE
STOMIIDAE
CHAULIODONTIDAE
Fig. 95. Weitzman's (1974) hypothesis of relationships of the sto-
miiform fishes. The Gonostomatidae and Stemoptychidae comprise the
Gonostomata and the remaining families comprise the Photichthya.
sister group to the Eurypterygii. Fink and Weitzman (1982)
agreed with this placement, provided more characters to sub-
stantiate it, and demonstrated monophyly of the stomiiforms.
Steyskal (1980) has presented arguments that the root of the
family-group names demands that these be altered from Sto-
miatidae and Stomiatiformes to Stomiidae and Stomiiformes,
respectively, and I use these forms throughout this paper.
As recognized by Weitzman (1974), there are two major sto-
miiform lineages, Gonostomata and Photichthya, both classified
at infraordinal rank, with families Gonostomatidae and Ster-
noptychidae in the former and families Photichthyidae, Sto-
miidae, Chauliodontidae, Astronesthidae, Melanostomiidae,
Malacosteidae, and Idiacanthidae in the latter (Fig. 95). I have
no disagreement with Weitzman's hypotheses of monophyly of
the Stemoptychidae, but our recent work on Diplophos (Fink
and Weitzman, 1982) caused us to question the monophyly of
the Gonostomatidae and Photichthyidae, and my work on the
barbelled stomiiforms, comprising the remaining families, has
cast doubt on the entire traditional arrangement of the included
26 genera as well as on the monophyly of the Photichthya. I
have found features which support new hypotheses of relation-
ship within the stomiiforms and will present some of these ideas
below. Some are more tentative than others. Weitzman is cur-
rently working on the genera he placed in the Gonostomatidae
and Photichthyidae.
First, I have found no evidence that Diplophos is the sister
group of any other genus of stomiiform and it may be, as Fink
and Weitzman (1982) suggested, the sister group of the rest of
the order. Specializations in the adductor muscles indicate that
Diplophos
GONOSTOMA
Cyclothone
Margrethia
BONAPARTIA
Triplophos
STERNOPTYCHIDAE
PHOTICHTHYA
Fig. 96. Hypotheses of stomiiforms as discussed herein. See text for
explanation.
Gonostoma. Cyclothone, Margrethia. and Bonapartia form a
monophyletic group, but what relationships within that group
are I cannot say, and presumably this will be treated by Weitz-
man. These hypotheses would cause a redefinition of the Gon-
ostomatidae, restricting it to the four genera mentioned just
above. Relationships of Triplophos are also unclear, and there
is evidence in the hyoid apparatus that it may be related to some
of the "photichthyans," rather than the gonostomatids, as
Weitzman ( 1 974) supposed. Weitzman ( 1974) established mon-
ophyly of the Stemoptychidae, and 1 have nothing to add to his
conclusions. Nevertheless, since he did not deal with monophyly
of the Gonostomatidae or with the sister group relationship of
the Stemoptychidae, there is no current evidence that the latter
is more closely related to some subset of the former, and I leave
that part of the phylogeny unresolved. These hypotheses are
summarized in Fig. 96. See also the paper by Ahlstrom, Rich-
ards, and Weitzman (this volume) on the Gonostomatidae, Ster-
noptychidae and other stomiiforms.
Within the "Photichthya," we have the same problem as with
the Gonostomatidae; that is, there is a diagnosable monophy-
letic unit (the barbelled forms) and an undiagnosed grade group,
the Photichthyidae.
My own efforts have been on the barbelled forms, currently
distributed in six families, as listed above. There have been no
strictly phylogenetic studies of relationships within the group,
but they were examined in a traditional sense by Parr (1927),
Regan and Trewavas ( 1 929, 1 930), and Beebe and Crane (1939).
FINK: STOMIIFORMS
183
My hypotheses are based on a study of 330 characters, mostly
taken from the skeleton, but with some from the head muscles,
photophores, and other parts of the soft anatomy. The conclu-
sions are presented in Fig. 97. Traditional families are not rec-
ognizable in this scheme of relationships.
Evidence for the arrangement of the genera is presented else-
where (Fink, in prep.), but some characters will be discussed
below, particularly those relevant to some of the larger portions
of the tree or in areas that might seem controversial to some
readers. For ease of communication, I will state here that my
choice of classification for this group is an expansion of the
traditional Stomiidae of Regan and Trewavas (see Fig. 97).
Monophyly of the Stomiidae is established on the basis of up
to 1 7 characters, including 1 ) presence of a mental barbel, 2) 5
hypurals in the caudal skeleton rather than 6, 3) lack of gill
rakers in adults, 4) a divided geniohyoideus muscle, and 5) a
portion of the adductor mandibulae inserting on the postorbital
photophore.
The Astronesthidae, as most recently discussed by Weitzman
(1967), consisted oi Astronesthes. Boroslomias. Heterophotus.
Neoncsthcs, and Rhadinesthes. As can be seen in Fig. 97, the
group is clearly not monophyletic. Neonesthes is the sister group
of all other stomiids, a hypothesis borne out by many characters
shared by the remaining stomiid genera, including lack of tooth-
plates on basibranchial 1, epibranchial 4, and on the posterior
edges of gill arches 1-4, and presence of rector muscles attaching
to the fifth ceratobranchial. The several equally parsimonious
constructions of stomiid relationships leave an unresolved tri-
chotomy at the next level, there being insufficient evidence re-
garding the positions of Aslronesthes. Boroslomias. and the re-
maining stomiids. This problem will be further discussed by
Fink (in prep.).
The remaining stomiids are united by such traits as lack of
toothplates on basibranchial 3 and position of the basihyal-
hypohyal ligament, as well as specializations of the dorsal and
anal fin skeletons. At this point there lies another unresolved
trichotomy, involving the groups Heterophotus plus Rhadi-
nesthes, Slomias plus Chauliodus, and the remaining stomiids.
Heterophotus and Rhadinesthes are documented as sister taxa
by several characters, including an elongate dorsal spine on the
cleithrum and a preopercle that is narrow at the area of the
symplectic-hyomandibular joint. That Chauliodus and Stomias
are sister taxa is supported by numerous characters, including
a nasal bone which forms a cup-like wall to the nasal capsule;
distribution of the palatine teeth into two areas, one anterior
and one well posterior; branchiostegals deeply bifurcated dor-
sally; and a distinct hexagonal pigment pattern in the skin. I do
not recognize the genus Macrostomias since work in progress
shows that those species are the sister group to a derived group
within Stomias.
The remaining genera, comprising the traditional families
Melanostomiidae. Malacosteidae, and Idiacanthidae, are united
by presence of many features, including no more than one pair
of toothplates associated with any basibranchial ossification,
and reduction of the distal radials of the pectoral fins.
As postulated by Regan and Trewavas (1930), I have also
found that Chirostomias and Tngonolampa are sister taxa based
on features such as fusion of the bilateral toothplates of basi-
branchials 2 and 3 and reduction of the supramaxiUa to a sliver
of bone. These genera are the sister group to the remaining
genera, a hypothesis supported by several characters, including
fewer than 6 branchiostegals articulating with the posterior cer-
Neonesthes
astronesthes
borostomias
Heterophotus
Rhadinesthes
Chauliodus
Stomias
Chirostomias
Trigonolampa
Thysanactis
leptostomias
Opostomias
Odontostomias
Flagellostomias
Photonectes
Echiostoma
Melanostomias
Idiacanthus
Tactostoma
Grammatostomias
Bathophilus
eustomias
Aristostomias
Malacosteus
Pachystomias
Photostohias
Fig. 97. Hypothesis of relationships within the Stomiidae, as dis-
cussed herein.
atohyal ossification, 3 or fewer distal pectoral fin radials, and
presence of a modification of the anterior pectoral fin rays into
a structure I call the "rod-ray complex."
For the remaining genera, I will concentrate on establishing
the major lineages as monophyletic and on areas that affect
traditional familial classifications of the group, particularly the
relationships of the "malacosteids" and Idiacanthus.
One monophyletic group is comprised of Flagellostomias.
Leptostomias. Odontostomias. Opostomias. and Thysanactis.
Among the diagnostic features are fusion of the distal cartila-
ginous tips of the lateral ethmoid and supraethmoid, and an
elongate opercular process of the hyomandibula.
The remaining genera are supported as monophyletic by nu-
merous characters, among them being lack of a retroarticular
(also lacking in Trigonolampa), and the form of the articulation
of the interhyal. The latter element articulates anterior to the
front margin of the cartilage between the hyomandibula and
symplectic and is bound to the metapterygoid by a ligament
from the anterior margin of the interhyal.
The Malacosteidae has traditionally been comprised of three
genera, Aristostomias. Malacosteus. and Photostomias, all of
which lack a floor to the mouth. The evidence shows that Pachy-
184
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
stomias also belongs to this group, and not with the other "me-
lanostomiids." This finding is not particularly radical, since
other authors have noted the close morphological resemblance
of that genus to the other three and indeed, it has been kept out
of the Malacosteidae mostly because the mouth floor is still
present, though thin, in members of the genus. The data are
insufficient to allow an unambiguous resolution of the interre-
lationships of these genera, but numerous characters support
the monophyly of the assemblage, including the suborbital pho-
tophore being ventral or posteroventral to the eye and the car-
tilage of the palatine arch being interrupted between the pos-
terior margin of the palatine and the rest of the arch.
Idiacanthus has usually been placed in a family by itself as
was done, for example, by Beebe (1934), primarily on the basis
of the specialized stalked-eyed larval stages and the degree of
sexual dimorphism. Beebe recognized that the genus was "closely
related to the Melanostomiatidae," as did Gibbs (1964b). Nei-
ther author suggested more precise relationships, and Beebe and
Crane (1939) showed Idiacanthus in a large multichotomy in
their figure of "relationships." Regan and Trewavas ( 1 930) con-
sidered Idiacanthus to belong with Melanostomias, Echiosto-
ma. and Photonectes. but did not say precisely where. My data
support placement of the genus as sister group to Tactostoma,
a genus described in 1 939. These two are then related to a group
of genera as shown in Fig. 97. Note that Melanostomias and
Echiostoma are excluded, being the sister group of the entire
assemblage. I am confident of the placement of Idiacanthus and
Tactostoma together, based on an array of characters, including
reduction of the basihyal to a thin, cylindrical element, origin
of the dorsal section of the medial division of the adductor
mandibulae muscle anterior to the insertion of the levator arcus
palatini muscle, and an extremely elongate body. But I am not
particularly confident in the placement of these two genera with
the others, even though the data appear impressive at first glance.
This lack of confidence is attributable to the fact that most of
those characters change at least three times in the entire tree,
leaving but one, lack of a posttemporal bone, as the only un-
reversed character supporting the hypothesis.
Another possibility is that Idiacanthus and Tactostoma are
the sister group of Melanostomias and Echiostoma. as suggested
in part by Regan and Trewavas (1930), apparently based on the
close morphological resemblance of Idiacanthus with the latter
two genera. Such a hypothesis would require some additional
reversals or independent losses, but as just noted, most of these
characters change several times even in the most parsimonious
tree. This part of the total phylogeny deserves more critical
examination, and it is hoped that larval specializations will be
found which will be found which will cause one hypothesis to
be clearly preferred over the other.
Regarding classification of the stomiiform fishes, it appears
that most of the traditional groups will cease to be recognized,
a move that was initiated by Weitzman (1974). A period of flux
should be expected until his curtent work is completed, but such
temporary instability is the current state of teleostean classifi-
cation at all levels, as phylogenetic methodology is applied with
increasing frequency. One might expect, however, that classi-
fication within the Stomiiformes will be stable sooner than that
in many other groups, because phylogenetic methods already
have been applied to it for several years. I will not present a
classification here, but I do provide such for the Stomiidae in
my revision of the group (Fink, in prep).
In summary, there is still much to be done in unravelling the
phylogenetic history of the main lineages of stomiiform fishes.
1 have outlined above areas where our knowledge is either in-
complete or poorly developed, and these should be the areas
where workers now concentrate their attention— to establish
monophyletic groups among the "primitive" stomiiforms and
to critically reexamine some of the hypotheses I have produced
within the barbelled stomiiforms. Some of this work is under-
way, using adult and sub-adult specimens, but the usefulness of
larvae is as yet unknown. The data presented in Ahlstrom's
(1974) work on patterns of metamorphosis in "gonostomatid"
fishes corroborate, when analyzed by phylogenetic methods, the
placement by Weitzman (1974) of many of those genera in an
expanded Stemoptychidae. An example of this is the presence
of photophores in clusters with common bases in those fishes
recognized by Weitzman as stemoptychids. Kawaguchi and
Moser (this volume) present the most comprehensive infor-
mation to date of stomiid larvae. Their data indicate that there
should be a plethora of characters for phylogenetic analysis and
that study of larvae should indeed prove useful in testing hy-
potheses of stomiid relationships. However, even a cursory ex-
amination of their data indicates that, as with characters in
adults, there appears to be a high degree of homoplasy. This is
an interesting phenomenon deserving further study.
Museum of Zoology, University of Michigan, Ann Arbor,
Michigan 48109.
Families Gonostomatidae, Stemoptychidae, and Associated Stomiiform Groups:
Development and Relationships
E. H. Ahlstrom, W. J. Richards and S. H. Weitzman
A summary of known information about the larvae and re- formation, both published and unpublished, gleaned from early
lationships of the stomiiforms with elongate gill rakers in life history stages and from adults. We also append some ten-
adults was published by Ahlstrom (1974). The present paper is tative new hypotheses of relationships within this "group" of
an addendum to that contribution and includes additional in- stomiiforms.
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
185
Table 47. Summary of Diagnostic Characters for Eggs of Certain Stomiiform Fishes.
Illus-
Species
Egg diameter
Oil globule
Diameter
Yolk
Special features
trated
Source
Argyropelecus
0.92-1.04
1
0.26-0.28
segmented
large oil globule
Yes
Sanzo, 1928
hemigymniis
Ichthyococcus
0.80
1
0.24
segmented
large oil globule
Yes
Sanzo, 1930b
ovalus
Maurolicus
1.63
1
0.25
segmented
hexagonal pattern
Yes
Mito, 1961a
muellen
1.32-1.58
1
0.26-0.28
segmented
on shell
Yes
Sanzo, 193 Id
1 inciguerna
0.58-0.74
none
irregularly
thin inner shell
Yes
Ahlstrom and
lucc'lia
segmented
membrane
Counts, 1958
powenae
0.75-0.85
1
0.17-0.19
segmented
no thin inner shell
membrane
Yes
Ahlstrom and
Counts, 1958
nimhana
0.64-0.72
none
irregularly
segmented
thin inner shell
membrane
No
Ahlstrom and
Counts, 1958
atlenuata
0.84-0.92
1
0.18-0.195
segmented
no thin inner shell
membrane
No
Sanzo, 193 Id
Gonosloma
0.80-0.81
1
0.20-0.21
—
—
No
Sanzo, 193 Id
denudatum
Ahlstrom (1974:672) favored recognition of one family for
those stomiiforms with elongate gill rakers in adults. According
to the rules of priority this would be the Stemoptychidae. Weitz-
man (1974:338) recognized three families, Gonostomalidae,
Photichthyidae, and Stemoptychidae, for the same stomiiforms,
the last family including the "maurolicin" genera formerly as-
signed to the Gonostomalidae and the deep-bodied stemop-
tychids traditionally assigned to the family. In a phylogenetic
or cladislic analysis this elongate gill raker bearing "group," if
recognized as a single family, is paraphyletic if one considers
certain of its subgroups as equivalent or higher taxonomic cat-
egories. For example, recognition of Ahlstrom's Stemoptychi-
dae, which would include the Stomiidae, a monophyletic group
with its members having a median barbel attached to the ventral
surface of the head in association with the hyoid bone and
lacking elongate gill rakers in adults, is incompatible with a
phylogenetic classification based on nested monophyletic groups,
since the Stomiidae is the sister group of another group within
Ahlstrom's Stemoptychidae. Furthermore, the character used
here to "define" the paraphyletic Stemoptychidae, the presence
of elongate gill rakers in adults, is excellent for use in a key for
identification purposes, but cannot be used as a synapomorphy
relating these fishes because it is primitive for stomiiforms.
Ahlstrom's Stemoptychidae is undefinable in a phylogenetic
analysis based on the information at hand. A resolution of the
use of familial and subordinal names in stomiiform fishes must
await completion of ongoing phylogenetic studies of these fishes.
Because these studies are incomplete, it is difficult to make
recommendations for names of certain stomiiform subgroups.
Among the stomiiforms with elongate gill rakers in adults, the
"family" problem is more complex than that recognized by
Ahlstrom (1974) or Weitzman (1974). We here recognize two
family names but these apply to only some of the 24 genera
listed below. We recognize the Stemoptychidae of Weitzman
(1974) and the Gonostomatidae in a new and restricted sense.
See discussion below.
The stomiiforms discussed here include the following 24 gen-
era, listed alphabetically, which have been variously recognized
as belonging to the families Gonostomatidae, Stemoptychidae,
Maurolicidae, and Photichthyidae:
Araiophos Grey (two species), Argyripnus Gilbert and Cramer
(four, possibly a few more), Argyropelecus Cocco (about sev-
en), Bonapartia Goode and Bean (one). Cyclolhone Goode
and Bean (twelve). Danaphos Bruun (one, possibly two), Dip-
lophos GUnther (two), Gonostoma Rafinesque (six), Ichthyo-
coccus Bonaparte (three), Manducus Goode and Bean (two),'
Margrethia Jespersen and Tuning (one, possibly two), Mau-
rolicus Cocco (one, possibly two). Photichthys Hutton (one),
Pollichthys Grey (one), Polyipnus Giinther (about sixteen),
Polymetme McCulloch (one, possibly four), Sonoda Grey (two),
Sternoptyx Hermann (two or three), Thorophos Bruun (two,
including Neophos Myers), Triplophos Brauer (one), Valen-
ciennellus Jordan and Evermann (one, possibly two), Vinci-
guerria Jordan and Evermann (five), Woodsia Grey (one),
and Yarella Goode and Bean (one).
' Grey (1964:88) recognized Manducus Goode and Bean, 1896 as a
junior synonym of Diptophos GxmXheT, 1873 because, as she stated ". . .
the differences appear to be of a specific rather than a generic nature
. . ." This was in the context of the kinds of differences Grey noted
separating other species of "gonostomatids." She did recognize both as
subgenera ot Diphphos. We recognize both as genera. The species were
most recently reviewed by Mukhacheva (1978) who recognized four
species, D. maderensis (Johnson), D. rebamsi Krefft and Farm, D. greyae
R. K. Johnson, and D. taenia Giinther. We have examined all four
species and find that D. taenia and D. rebamsi have the cartilages of
the two medial proximal pectoral radials, radials III and IV in the
terminology of Fink and Weitzman (1982:66), fused while retaining two
bony elements separate as reported for D. taenia by Fink and Weitzman
(1982:65-67). Furthermore, one of the distal radials is out of line, not
in a single series in these two species. These characters are specialized
for these species. In Manducus maderensis and A/, greyae there are four
completely distinct proximal radials and the distal radials are all in a
simple straight series. Because the pectoral radial morphology in Diplo-
phos taenia and D. rebamsi may be an intermediate stage of a transition
series between radials such as are found in Manducus maderensis and
M. greyae and those in the "photichthyid" genera, we recognize Man-
ducus &% a genus and apparent sister group of the "photichthyid" genera
as well as the Stomiidae, nearly all of which have the radials 111 and IV
completely fused to one bone. A few stomiids have an apparent neo-
morph condition in which the third proximal radial is divided into two
radials, giving a total of four proximal radials. See also text discussion.
186
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 48. Summary of Meristic Characters for Adults of Certain Stomiiform Fishes.
No.
species
Fin rays
Branchi-
oslegal
rays
No. of
vertebrae
Genera
Dorsal
Anal
Pectoral
Pelvic
No. of gill rakers
Araiophos
2
13-20
20-29
16-18
5
9-11
43-45
2-3 + 12-19 = 14-22
Argyripnus
4+
11-12
11-15 + 8-12 = 22-29
15-19
6-7
8-10
41-46
4-7 + 12-19= 16-26
Argyropelecus
7
(8)9(10)
6-8 + 5-6=11
-13
10-11
7
7
34-40
15-24
Bonapartia
1
17-20
29-31
14-16
7-8
13-16
37
5-6 + 11-12= 16-18
Cydothone
12 +
12-15
16-21
9-13
6-7
10-14
29-33
4-10 + 9-18= 14-27
Danaphos
1
6
24-25
13-14
6
9-10
38
2 + 11-13 = 13-15
Diplophos
2
10-13
47-69
8-9
7
10-14
44-94
3 + 7-9= 10-12
Gonostoina
6
10-18
21-31
9-13
6-8
10-13
37-40
5-11 + 10-17= 15-27
Ichlhyococcus
3
10-15
13-17
7-8
6-7
11-12
38^7
7-11 + 15-26 = 22-37
Manducus
2
11-13
36-59
9-11
8
11-14
63-76
3-5 + 8-10 = 12-14
Margrethia
1
15-16
21-26
13-15
8
13
34
5 + 10-11 = 15-16
Mauro/icus
1
9-12
8-10 +
11-15 =
19-27
17-20
6-7
9-10
33-35
4-8 + 17-22 = 22-30
Photkhthys
1
12-13
23-26
9
6-7
20-21
51
4-5 + 11 = 15-16
Pollichlhys
1
10-12
22-30
8
6-7
11-12
40
4-5 + 11-12= 15-17
Polyipnus
17
10-17
13-19
12-16
7
9
31-36
10-28
Polymetme
3
11-13
24-33
9-11
7 (8?)
12-14
44-45
5-8 + 9-12= 15-19
Sonoda
2
8-9
8-10 +
14-16 =
22-25
13-15
6
8-10
40?
3-5 + 15-18= 18-21
Sternoptyx
3
8-11
14-16
10-11
7
7
28-31
7-9
Triplophos
1
10-12
53-63
9-11
6-7
11-14
ca 60
9 + 14-16 = 23-25
Thorophos
2
8
38
13
7
7-8
40-45
5 + 13-14= 18-19
Vatenciennellus
2 or 3
7-12
22-25
12-13
6-9
9-10
32-33?
2-3 +12= 14-15
Vinciguerha
4
13-16
12-17
9-10
7
10-12
38-42
3 + 11-23-11 = 15-33
Woodsia
1
11-12
14
9-10
7-8
17
42-45
3-5 + 13 = 16-18
Yarella
■>
14-16(17)
(28)29-
U
8-10
6-7
13-16
45-54
6-7 + 12-16= 18-22
Table 49. Position of the Dorsal and Anal Fin and Condition of the Adipose Fin in Certain Stomiiform Fishes.
Dorsal &n position
Genus
Adipose fin
Anal origin in advance of dorsal fin. Dorsal origin
opposite 5th or 6th anal ray
Anal origin opposite dorsal origin
Anal origin opposite last dorsal fin ray
Anal origin well in advance of dorsal by 9 rays
Anal origin opposite dorsal fin or slightly behind
Anal origin behind dorsal fin
Anal origin beneath 5th ray or behind dorsal
fin
Anal origin opposite or 3-4 rays in advance of
dorsal origin
Anal origin behind dorsal fin by a space = '/2 dor-
sal base
Anal origin beneath 3rd from last or last dorsal
fin ray
Anal origin beneath 5th dorsal fin ray
Anal origin beneath last dorsal fin ray
Anal origin behind dorsal fin
Anal origin beneath 3rd dorsal fin ray
Anal origin usually beneath middle of dorsal fin
Anal origin beneath end of dorsal fin
Anal origin in advance of dorsal. Dorsal origin
above 5th anal ray
Anal ongin opposite dorsal origin
Anal origin beneath end of dorsal fin
Anal origin in advance of dorsal origin by 3 or 4
rays
Anal origin 1 or 2 rays in advance of dorsal origin
Anal ongin beneath middle of dorsal fin
Anal origin behind middle of dorsal fin by dis-
tance about = dorsal base
Anal origin beneath middle of dorsal fin
Anal opposite dorsal at 8 mm, adult
position at 1 1 m
Anal origin opposite dorsal origin
Anal origin behind dorsal fin
Same as adult
Same as adult
Same as adult
Anal origin beneath end of or behind
dorsal fin
Same as adult
Anal origin behind dorsal fin
Unknown
Araiophos
Argyripnus
Argyropelecus
Bonapartia
Cydothone
Danaphos
Diplophos
Gonostoma
Ichthyococcus
Manducus
Margrethia
Maurolicus
Photichthys
Pollichlhys
Polyipnus
Polymetme
Sonoda
Sternoptyx
Triplophos
Thorophos
Valenciennellm
Vinciguerria
Woodsia
Yarella
Present or ab-
sent
Present
Present or ab-
sent
Absent
Absent
Absent
Absent
Present or ab-
sent
Present
Absent
Same as adult
Present
Anal origin beneath middle of dorsal
Present
fin, advances to adult condition as
juveniles
Unknown
Present
Anal origin advances forward beneath
Present
dorsal fin
Same as adult
Present or ab
sent
Unknown
Present
Unknown
Absent
Anal origin behind dorsal fin
Present
Unknown
Absent
Unknown
Present or ab
sent
Same as adult
Present
Same as adult
Present
Same as adult
Present
Same as adult
Absent
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
187
Table 50. Dernition of Alphabetical Symbols used for Designating Photophores in Deep Bodied Sternoptychids and Other Stomiiform
Fishes.
Other slomiiforms
Deep bodied slemoplychids
SO Symphyseal photophores (organs) located at tip of
lower jaw.
Orb Photophores associated with the eye located ante-
rior and posterior of orbit.
Op Photophores on opercle series generally three, cod-
ed as follows 1/(1 -I- 1).
Br(BRP) Photophores located on the branchiostegal mem-
branes.
Is(I) Photophores located on the isthmus.
IP Photophores of the ventral series found from the
isthmus to the base of the pectoral fin.
PV Photophores of the ventral series found from the
pectoral fin base to the pelvic (ventral) fin base.
VAV Photophores of the ventral series found from the
pelvic (ventral) fin base to the anal fin base.
AC Photophores of the ventral senes found from the
anal fin base to caudal fin base of the ventral se-
ries.
IC Summary of photophores of the ventral series from
the isthmus to caudal fin base
(IP + PV + VAV + AC).
IV Summary of photophores of the ventral series from
isthmus to pelvic (ventral) fin base (IP + PV).
OV Photophores of the lateral series from the opercle
to pelvic (ventral) fin base.
VA(VALA) Photophores of the lateral series from the pelvic
(ventral) fin base to the anal fin base.
OAA Summary of photophores of OV plus VA series.
OA(OAB) Summary of lateral photophores from the opercle
to anal fin base (OV + VA).
OAC(OC) Entire lateral series on body sides just dorsal to
ventral series and extending from opercular
border, or just medial to it, over anal fin to cau-
dal fin base.
ODM Photophores (organs) found dorsal to the lateral
midline (found only in Gonosloma gracile).
SO
PO
PTO
PRO
Br
Is
AB
PAN
AN
SC
SAB
SP
L
SAN
Subopercle photophore which is equivalent to pos-
teriomost photophore in opercular series of gon-
ostomatids.
Photophore located anterior to orbit.
Photophore located posterior to orbit and may be
equivalent to upper photophore of opercular se-
ries of gonostomatids.
Preopercular photophore, used for an PO photo-
phore dorsal to ventral limb or preopercle.
Same as gonostomatid definition.
Same as gonostomatid definition.
Photophores of ventral series located abdominally
between pectoral fin base and pelvic fin base and
equivalent to PV in gonostomatids, plus a few
posterior photophores of the IP series.
Photophores found anterior to anal fin and may be
equivalent to VAV or VA in gonostomatids.
Photophores found above anal fin.
Photophores found on lower (sub) caudal peduncle.
Together with AN group may be equivalent to
AC in gonostomatids.
Photophores located above (supra) to the abdomi-
nal series and may be equivalent to VA in gon-
ostomatids.
Photophores located above (supra) the pectoral fin
and may be equivalent to OV in gonostomatids.
Photophore located laterally above PAN (found
only in Polyipnus).
Photophores located above (supra) to anal photo-
phores and equivalent to part of AC series.
Some genera are extremely rare (i.e., Thorophos and Sonoda)
while Others represent the most abundant vertebrate animals
on earth (Cyclothone and I'incigiierria).
Developmental information has been published for 16 of these
genera (12 prior to Ahlstrom, 1974; 3 by Ahlstrom, 1974; and
one by Ozawa, 1976).
Development
Eggs.— Eggs were desciibed for Argyropelecus hemigymnus by
Sanzo (1928); for Ichthyococcus ovatus by Sanzo (1930b); for
Maurolicus muelleri by Sanzo (193 Id), Mito (1961a). and Oki-
yama (1971); for Vinciguerna lucetia. V. poweriae. and I', nim-
baria by Ahlstrom and Counts ( 1 958); for V. attenuata by Sanzo
(193 Id); and for Gonostomadenudatumby Sanzo (\9'i\d). Oth-
er accounts provide minimal details of ovarian eggs of other
species. The details of egg characters are summarized in Table
47.
Larvae. — Much has been accomplished for the identification of
the larvae of these stomiiform genera and now descriptions are
available for all except Manducus. Triplophos, Polymetme, Pho-
tichthys, Thorophos, and Sonoda. The larvae tentatively iden-
tified as Polymetme by Ahlstrom ( 1 974), on further examination
by one of us (Richards), were determined to be Pollichthys. One
stomiiform larval form has been described but not assigned to
a genus [designated "Maurolicine Alpha" by Ahlstrom (1974:
670)]. It presumably is the larva of some stemoptychid (as de-
fined by Weitzman, 1974). Descriptive details and illustrations
of several species were given by Ahlstrom (1974). Here we pro-
vide new or additional data including characters useful in iden-
tifying these larvae and illustrations of all the species described
to date, including some illustrated for the first time.
The identification of stomiiform larvae with elongate gill rak-
ers as adults requires a knowledge of developmental data from
larvae, juveniles, and data from adults of the following char-
acters: counts of fin rays, teeth, and other meristic characters as
photophores; patterns of photophore development; and distri-
butions (patterns) of dark chromatophores (dark pigment cells).
With those sets of data, nearly all species should be identifiable
at least to genus, and in cases of complete data, to species. A
188
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 51. Photophore DrsTRiBUTiON in Certain Stomtiform Genera. Refer to text and Table 50 for definition of codes.
Photo-
phores in
No. of
group of
rows
so
orb
OP
BR
IS
IV
VAV
AC
glands
Araiophos
1
No
1
1
5-7
Yes
(2)
+ (3) + 3-4 + (2)
= 10-11
3-5
6-8
Yes
Argyripnus
2
No
1
3
6
Yes
(6) + (10)
(18-28)
(4-5) + (12-18)
= 35-51
Yes
Argyropelecus
2
No
2
2
6
Yes
18
4
10
Yes
Bonapartia
1
Yes
1
3
11-13
No
14-16
5-6
18-20
No
Cydothone
2
No
1
2
8-11
No
12-14
4-5
12-16
No
Danaphos
2
No
1
2-3
6
Yes
18
5
22-26
Yes
Diplophus
3 +
Yes
1
3
7-12 +
0-3
9
Yes
33-*9
13-17
33^9
No
Gonostoma
2
Yes
1
2-3
No
11-16
3-10
15-23
No
Ichthyococcus
2
No
2
3
11-12
Yes
25-28
9-14
12-14
No
Manducus
2 +
Yes
1
3
8-13
Yes
30-33
12-14
28-39
No
Margrethia
1
No
1
3
9-12
No
13-15
4
17
No
Maurolicus
2
Yes
1
3
(6)
Yes
(6) + (12-13)
= 18-19
(6)
1 + (14-18) + (7-
= 22-27
9)
Yes
Pholichlhys
2
Yes
2
3
17-18
Yes
10
+ 14-15 = 24-25
15-17
16-18
Yes
Poltichthys
2
Yes
2
3
8
Yes
21-23
7-9
18-21
No
Polyipnus
2
No
2
2
6
Yes
16
5
10-18
Yes
Polymetme
2
Yes
1
3
9-10
Yes
19-21
7-8
21-25
No
Sonoda
2
No
1
3
6-7
Yes
6 + 10= 16
7-8
(16-21) + (19-24)
(5-6) + (5-6) + (5-
= 36-43
or
-6)
Yes
Siernoptyx
2
No
2
2
3
Yes
15
3
7
Yes
Tnplophos
2 + 3 or 4
Yes
1
3
8-13
Yes
24-30
5-7
35-41
No
Thorophos
2
Yes & no
1
3
6
Yes
17
5
13-15
Yes
Valenciennellus
2
No
1
3
6
Yes
(3
+ (4) + (16-17)
= 23-24
(4H5)
3-6 or 9-17
Yes
Vincignerna
2
Yes or no
2
3
7-9
Yes
21-24
7-11
12-15
No
Woodsia
2
Yes
2
3
14
Yes
25
11-12
12
No
Yarella
2+ sev
Yes
1
3
11-13
Yes
23-25
9-12
20-28
No
summary of several meristic characters for genera is given in
Table 48. The position of the dorsal and anal fins is also a helpful
aid, but caution must be used since their positions relative to
other structures may change with growth. Also, the presence or
absence of the adipose fin is helpful, but again, caution is in
order because this fin is fragile and often damaged or lost due
to contact with a net. These fin features are given in Table 49.
Of special importance in identifying lai^ae and adults is the
distribution and patterns of the photophores. This includes the
number in each series, the patterns of their distribution in re-
lation to each other, and especially the sequence of development
which Ahlstrom (1974) stressed. Some confusion appears in the
literature because more than one alphanumeric code has been
developed to indicate, in some cases, the same sets of photo-
phores in different stomiiform groups. A further complication
is that the deep-bodied stemoptychids have a different code
because of their altered body shape as adults and homologies
were considered uncertain. Weitzman (1974:461), because he
united the "maurolicin" and deep-bodied stomiiforms as one
family considered the different termmologies "artificial" and as
obscuring homologies. He therefore discussed and presented a
synonymy of stomiiform photophores. We have defined the
alphabetical codes in Table 50 and included what we believe
are equivalent photophores in stomiiforms. In this code, par-
enthetical numbers indicate photophores found in common
glands whereas non-parenthetical numbers indicate that the
photophores are single. The distribution of photophores for each
genus is given in Table 51. Table 52 provides sequences of
photophore formation for Bonapartia, Margrethia, and Gon-
ostoma. Table 53 provides similar information for Araiophos.
Maurolicus, Danaphos, Valenciennellus. and Argyripnus; while
Table 54 provides similar data for Polyipnus. Argyropelecus,
and Siernoptyx. Diagnostic pigmentation and morphometric
characters are summarized in Table 55. Illustrations (Figs. 98
to 104) are provided for the genera for which larvae are known
and for many of the known species. In addition, the following
authors provide specific information which will aid in larval
identifications: Jespersen and TSnmg (1919, 1926), Sanzo
(193 Id), Ahlstrom and Counts (1958), Ahlstrom and Moser
(1969), Ozawa (1976), Grey ( 1 964), Badcock and Merrett (1972),
Kawaguchi and Marumo (1967), Okiyama (1971), Badcock
( 1 982), Rudometkina (1981), Gorbunova (1981), Mukhacheva
(1964), and Ahlstrom (1974).
Relationships
There has been a dichotomy of opinions about the interre-
lationships of the genera and the use of family names based on
the use of larval versus adult morphological characters. Ahl-
strom (1974:670-672) presented his views on this group based
on larval characters, principally the mode of photophore for-
mation. The suggested relationships resulting from his analysis
contrasted in part with those of Weitzman (1974:472), whose
views were based on study of adult osteology and soft anatomy.
Both Ahlstrom and Weitzman in addition to their own data,
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
189
Table 52. Sequence of Photophore Formation in Bonapania. Margrethia. and Gonostoma.
BR
pv
VAV
AC
Bonapartia
pedaliota
Margrclhia
obtusirostra
Gonostoma
etongatum
Gonostoma
demidalum
Gonostoma
gracile
Gonostoma
ehelingi
Gonostoma
bathyphilum
Gonostoma
allanlicum
adult
9.5
11.5
12.0
14.0
15.0
16.0
23.0
adult
5.8
6.4
8.0
11.3
15.0
adult
6.0
7.5
7.9
10.2
13.0
14.0
16.7
22.5
adult
18.25
19.0
20.75
24.75
29.65
34.0
39.0
adult
15.5-5-17.0
20.0
22.0
adult
13.8
15.0
adult
11.0
14.8
adult
12.0
13.0
14.5
17.8
18.8
23.8
I
0
0
0
0
0
1
1
1
1
0
0
0
0
0
I
1
I
0
1
1
1
0
0
1
0
1
1
0
0
0
1
1
1
I
0
0
0
0
0
0
0
0
0
0
0
0
0
I
0
0
0
0
0
0
n
I
1
0
0
0
0
0
0
I
1
0
1
1
1
0
0
0
0
0
1
0
0
0
0
0
0
11-13
2
3
4
5
5
6
II
9-12
0
0
2
6
9
9
0
0
0
2/1
2
2
3
9
9
0
0
I
3
5
9
9
9
0
2
9
9
0
0
9
0
4
9
0
0
0
4
9
9
14-15
3
5
5
10
9
II
14
13-15
2
6
10
14
14
15
0
5
4
10
11
II
II
15
15-16
1
2
3
6
14
16
16
13-15
0
13
14
15
7
9
11-12
5
10
15-16
0
1
2
13
16
16
5-<6)
0
2
2
4
3
5
5
4
0
2
4
4
4
(4)-5
0
0
0
2
3
2/3
4
5
5
0
0
1
3
5
5
5
4-5
0
5
4
10
0
0
4-5
0
2
5
0
0
0
3
5
5
16-18 + 2-3
0
0
0
3 + 1
I + 1
5 + 2
14 + 2
13-14 + 3-4
0
1 + 2
1 + 2
5 + 3
11+4
21-23
0
0
0
0
0
1 +
1 +
22
17-20
0
+ 2
+ 3
3 + 3
11+3
15 + 5
15 + 5
17-19
0
17
18
19
0
0
20-21
0
0
19
0
0
0
1
19
19
II-
0 Grey, 1964
0 Grey, 1964
0 Original
0 Grey, 1964
0 Grey, 1964
0 Original
0 Jespersen and TSning, 1919
0 Grey, 1964
0 Grey, 1964
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Ahlstrom, 1974
13-15 Grey, 1964
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Original
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Grey, 1964
0 Jespersen and Tuning, 1919
13 Grey, 1964
13-15 Grey, 1964
0 Sanzo, 1912b
0 Sanzo, 1912b
0 Sanzo, 1912b
0 Sanzo, 1912b
0 Sanzo, 1912b
13 Sanzo, 1912b
13 Sanzo, 1912b
1 2 + 6-7 Kawaguchi and Marumo, 1 967
0 Kawaguchi and Marumo, 1967
0 Kawaguchi and Marumo, 1967
12 + 4 Kawaguchi and Marumo, 1967
21 Grey, 1964
0 Ahlstrom, 1974
0 Ahlstrom, 1974
14 Grey, 1964
0 Ahlstrom, 1974
0 Ahlstrom, 1974
13 Grey, 1964
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Ahlstrom, 1974
0 Original
0 Ahlstrom, 1974
13 Original
used the results of photophore anatomy research by Bassot ( 1 966,
1971) to support their conclusions. These results seemingly
completely supported Weitzman's referral of genera to family
groups and agreed with Ahlstrom except for placement of three
genera — C'lr/or/iowc, Diplophos (including Mandncus), and
Tnplophos.
One of us (Weitzman), continues to study relationships of the
stomiiforms with elongate gill rakers in adults and we offer the
following analysis as a current comment on the status of our
knowledge of these fishes. The two concepts of relationships by
Ahlstrom and Weitzman may be compared as follows: Ahlstrom
(1974:670-672) stressed relationships of taxa based on photo-
phore patterns and development. Ahlstrom (1974:672) consid-
ered the stomiiforms with elongate gill rakers in adults as a
group comprised of three groups of genera, with any subdivision
being into two subfamilies based on photophores occurring in-
dividually or in clustered groups. These groups of genera in-
clude: (I) Those with individual separate photophores, most of
the photophores developing simultaneously and initiated as a
"white" photophore stage. This group includes Manducus, Dip-
lophos, Cyclothone. Yarella, Pollichthys, V'inciguerha. Wood-
sia, Ichlhyococcus. and presumably Triplophos and Polymetme,
190
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
Table 53.
Sequence of
Photophore Formation in Araiophos, Maurolicus, Danaphos, Valenciennellus, and Argyripnus.
ORB
OP
so
BR
IP
PV
VAV
AC
OA
Source
Araiophos
adult
1
1
0
(6)
(2)
(3) + 3-
(3)
(2) + 2 + (2)
No
Ahlstrotn and Moser,
1969
eastropas
4 + (2)
11.2
0
0
0
(3)
0
(2)
0
0
-
Ahlstrom and Moser,
1969
Maurolicus
adult
1
3
1
(6)
(6)
(12)
(6)
3/(4) + (8)
(2) + 7
Ahlstrom,
1974
muelleri
5.5
0
0
0
(1/2)
0
0
0
0
0
Ahlstrom,
1974
6.2
0
0
0
(2)
0
(2)
0
0
0
Ahlstrom,
1974
6.5
0
0
0
(2)
0
(4)
0
0
0
Ahlstrom,
1974
6.7
0
1
0
(3)
0
(5)
0
0
0
Ahlstrom,
1974
6.9
1
0
(4)
0
(8)
0
0
0
Ahlstrom,
1974
7.5
1
0
(4)
I
(9)
0
0 + (2) + 0
0
Ahlstrom,
1974
8.6
2
0
(5)
(3)
(12)
(2)
0 + (3) + (3)
0
Ahlstrom,
1974
9.0
2
0
(5)
(3)
(11)
(2)
0 + (3) + (3)
1
Ahlstrom,
1974
9.7
3
0
(5)
(5)
(11)
(3)
0 + (4) + (6)
(2)+ 1
Ahlstrom,
1974
10.8
3
0
(6)
(5)
(12)
(4)
0 + (5) + (6)
(2) + 2
Ahlstrom,
1974
13.5
3
0
(6)
(6)
(12)
(6)
0 + (9) + (7)
(2) + 6
Ahlstrom,
1974
Danaphos
adult
1
3
0
(6)
(3) + (4)
(11)
(5)
(3) + 16 +
6
Ahlstrom,
1974
oculatus
(4)+ 1
Ahlstrom,
1974
16.5
0
0
0
(2)
0
0
0
0
0
Ahlstrom,
1974
16.5
0
0
0
(3)
0
(3)
0
0
0
Ahlstrom,
1974
19.2
0
0
0
(4)
0
(10)
0
0
0
Ahlstrom,
1974
21.0
1
1
0
(5)
(2) + (4)
(10/11)
0
(2) + 0 + 0 + 0
0
Ahlstrom,
1974
21.3
1
1
0
(4/5)
(3) + (4)
(10)
0
(3) + 0 + (2) + 0
0
Ahlstrom,
1974
21.8
I
2
0
(5)
(3) + (4)
(11)
(2)
(3) + 8 + (4) + 0
2
Ahlstrom,
1974
24.2
I
2
0
(6)
(3) + (4)
(11)
(2)
(3) + 9 + (4) + 0
2
Ahlstrom,
1974
Valenannellus
adult
1
3
0
(6)
(3) + (4)
(16-17)
(4-5)
(3) + (3) + (3) +
(2) + 3
Ahlstrom,
1974
thpunculatus
(2) + (4)
7.8
0
0
0
0
0
0
0
0
0
Original
8.6
0
0
0
(3)
0
(3)
0
0
0
Ahlstrom,
1974
9.5
0
0
0
(4)
0
(6)
0
0
0
Ahlstrom,
1974
11.0
0
0
0
(4)
0
(10)
0
0
0
Original
12.0
0
0
0
(4)
0
(13)
(2)
0
0
Ahlstrom,
1974
12.4
1
0
0
(5)
0
(15)
(2)
0
0
Original
13.0
1
0
0
(5)
(2)
(15)
(2)
0
0
Original
13.2
0
0
0
(4)
0
(14)
(3)
0
0
Ahlstrom,
1974
14.0
1
0
0
(5)
(4)
(15)
(5)
0
0
Original
17.0
1
2
0
(4-5)
(3) + (4)
(15)
(5)
(3) + (3) + 0 +
(3) + (4)
(2)
Grey, ige-)
Argyripmis
adult
1
3
0
(6)
(6)
(10)
(26)
(5) + (17)
(3) + 4
Badcock and Merrett,
1972
atlamicus
18.7
1
2
0
(6)
(3)
(10)
(3)
(4) + (4)
0
Badcock and Merrett,
1972
16.8
1
2
0
(6)
(3)
(10)
(2)
(4) + (3)
0
Badcock and Merrett,
1972
although their development is not known. (2) Those with in-
dividual, separate photophores that have a gradual, protracted
metamorphosis. This group includes Bonapartia, Margrelhia.
and Gonostoma. (3) Those with some individual photophores
but some or most of the photophores with common bases [ac-
tually a common lumen, during development at least] and hav-
ing a gradual, protracted metamorphosis. This group includes
Araiophos, Maurolicus, Danaphos, Valenciennellus, .Argyrip-
nus, Polyipnus, Argyropelecus, Sternoptyx, and presumably
Thorophos and Sonoda ahhough their development is unknown.
Groups (1) and (2) comprised the subfamily Gonostomatinae
and Group (3) comprised the Stemoptychinae in Ahlstrom's
concept. Group (3) is equivalent to Weitzman's Stemoptychi-
dae. The genus Gonostoma was considered "pivotal" by Ahl-
strom; that is, its relationships could be with either the gonos-
tomatines or the stemoptychines of his concept. In Ahlstrom's
conclusions, the photophore pattern of Group (1) is most like
that of the stomiid groups discussed by Fink in this volume.
Weitzman's classification (1974) concentrated in most detail
on a hypothesis of phylogenetic relationships within the family
Stemoptychidae as he defined it. Weitzman (1974) pointed out
that more detailed studies should be conducted on other sto-
miiform genera in the future, but he did discuss their possible
relationships. Based on the number of proximal pectoral-fin
radials, he established two infraorders for stomiiform fishes.
Members of the Infraorder Gonostomata were considered to
have four proximal pectoral-fin radials (except Cyclothone with
one). This infraorder was divided into two families based prin-
cipally on Bassot's photophore findings; Family Gonostomati-
dae with Beta type photophores comprised of Diplophos in-
cluding Manducus), Triplophos, Bonapartia, Margrethia.
Gonostoma, and Cyclothone and the Family Stemoptychidae
with Alpha type photophores comprised of Thorophos, Araio-
phos, Maurolicus, Danaphos, Valenciennellus, Argyripnus, Son-
oda, Polyipnus, Argyropelecus, and Sternoptyx. The problem
with Weitzman's Gonostomata is that it was based on a prim-
itive character for the stomiiforms, four pectoral-fin radials, and
this character cannot be used as a synapomorphy to define a
subgroup of stomiiforms. The non-stemoptychid and non-gon-
ostomatid genera, along with the stomiiform families possessing
barbels originating from the hyoid bone and lacking elongate
gill rakers in the adults (the Stomiidae of Fink, this volume).
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
191
Table 54. Sequence of Photophore Formation in Polyipnvs. Arcyropelecvs and Sternoptyx.
OP
PRO +
Size
PC
PTO
BR
IS
SO
SP
AB
SAB
PAN
AN
SAN
LSC
Source
Polyipnus polli
adult
1
1
6
6
1 + 1
3
10
3
5
8
3
14
Baird, 1971
4.3
0
0
2
0
0+ 1
0
1
0
0
0
0
0
Original
4.8
1
0
4
2
1 + 1
0
3
0
0
0
0
0
Original
5.5
1
1
6
4
1 + 1
2
8
0
0
0
0
0
Original
6.0
1
1
6
6
1 + 1
2
10
0
1
0
0
0
Original
7.5
1
1
6
6
1 + 1
2
10
0
3
0
0
0
Original
9.0
1
1
6
6
1 + 1
2
10
0
3
2
0
2
Original
9.6
1
1
6
6
1 + 1
2
10
1
3
2
0
2
Original
13.5
1
1
6
6
1 + 1
3
10
3
5
4
0
14
Original
15.3
1
1
6
6
1 + 1
3
10
3
5
4
1
14
Original
17.0
1
1
6
6
1 + 1
3
10
3
5
4
2
14
Original
18.4
1
1
6
6
1 + 1
3
10
3
5
6
3
14
Onginal
23.5
1
1
6
6
1 + 1
3
10
3
5
7
3
14
Original
Argyropelecus
adult
1
1
6
6
1 + 1
2
12
6
4
6
0
4
Baird, 1971
hemigymnus
10.92
0
0
4
6
0 + 1
0
7
0
0
1
0
2
Sanzo, 193 Id
9.92
0
0
6
6
0+ 1
0
9
0
0
2
0
3
Sanzo. 193 Id
7.84
1
0
6
6
1 + 1
2
12
0
0
3
0
4
Sanzo. 193 Id
11.20
1
1
6
6
1 + 1
2
12
2
3
4
0
4
Sanzo, 193 Id
Arygropelecus sp.
adult
1
1
6
6
1 + 1
2
12
6
4
6
0
4
Baird, 1971
4.5
0
0
0
0
0 + 0
0
0
0
0
0
0
0
Original
9.5
0
0
6
6
0 + 1
0
6
0
0
1
0
0
Original
9.5
0
0
6
6
0 + 1
0
8
0
0
3
0
0
Original
7.0
1
0
6
6
1 + 1
2
12
0
0
3
0
3
Original
7.0
1
0
6
6
1 + 1
2
10
0
0
3
0
4
Original
7.4
1
1
6
6
1 + 1
2
12
4
4
4
0
3
Original
10.0
1
1
6
6
1 + 1
2
12
5
4
5
0
4
Original
Sternoplyx sp.
adult
1
1
3
5
1 + 1
3
10
0
3
3
I
4
Baird, 1971
4.8
0
0
0
0
0 + 1
0
0
0
0
0
0
0
Original
7.5
0
0
0
0
0 + 1
0
0
0
0
0
0
0
Original
7.8
0
1
2
3
0 + 1
0
4
0
0
0
0
0
Onginal
8.1
0
1
2
4
0 + 1
2
7
0
0
0
0
0
Original
7.6
0
1
2
5
0+ 1
3
10
0
1
3
0
1
Original
Table 55. Diagnostic Pigment Characters and Unusual Morphometric Features of Some Stomiiform Larvae.
Genus/species
Diagnostic character
Diplophos taenia
Bonapartia pedaliota
Margrelhia obtusirostre
Gonostoma
Cyclolhone
Yarella blackfordi
Pollichthys mauli
Vincigucrria
Woodsia nonsuchae
Ichlhvococcus ovatus
Pigment spots on dorsal and ventral midline. Extremely elongated larvae.
Similar to Gonostoma but lacks deep pigment spot behind eyes and has pigment on medial portion of caudal
Ijeduncle.
A distinct vertical streak of pigment on caudal peduncle in most specimens.
All species usually have deep pigment spot behind eyes. Specific differences among the species are as follows:
G. elongatum. G. gracile and G. ehelingi lack pigment on caudal peduncle; G. bathyphilum has pigment
spots on dorsal edge of caudal peduncle; G. atlanticum has pigment over medial portion of caudal peduncle
(closely resembles Cyclothone in ventral pigmentation and swimbladder position); G. denudatum has broad
streak of pigment diagonally over caudal fin base from dorsal caudal peduncle to base of lower caudal fin
rays.
A distinct, dark streak or intense melanophore over and parallel to the parhypural on the caudal fin base,
pigmentation over gut and along ventral margin of tail and a conspicuous swimbladder.
Myosepta pigmented over caudal peduncle giving chevron appearance.
No pigment except for the eyes. Very similar to Vinciguerria in other aspects.
All species have medial or ventral margin caudal pigment spot. I '. nimbaria and V. lucetia have the caudal
pigment spot restncted to the ventral margin of the caudal fin base and pigment above the anal fin. V.
attenuata and I '. poweriae has the caudal pigment spot in a medial position and no pigment above the anal
fin. r. attentuata has pigment over the airbladder which is lacking in C. poweriae. V. poweriae has a struc-
ture above the anal papilla which may appear as pigment. V. mabahiss is similar to V. nimbaria and is
restricted to the Red Sea (Johnson and Feltes, 1984).
Melanophores profusely distributed on all myomeres below the lateral midline. Broad pigment band along
roof of mouth continuous with trunk pigment. Also has a trailing gut and elongated rays on pectoral fin,
both of which may be missing.
Pigment profusely distributed on all myomeres below the lateral midline. Elongate pectoral fin rays and a
trailing gut, both of which may be missing.
192
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
Fig. 98. Lateral views from top to bottom: Diplophos taenia 22.0 mm SL, R/V OREGON II Cr. 126, Sta. 36754, 27''30'N, 092°30'W, May
10, 1982, drawn by J. C. Javech; I'lncigiierria lucelia 9.0 mm SL modified after Ahlstrom and Counts (1958); I'lnciguerna powcnae 1 1.5 mm
SL, R/V OREGON II Cr. 126, Sta. 36746, 27°59.9'N, 088°00'W, May 8, 1982, drawn by J. C. Javech; and linciguerna atlenuata 9.7 mm SL
modified after Jespersen and Tuning (1926).
Fig. 99. Lateral views from top to bottom: Pollichlhvs niauli 14.5 mm SL, R/V OREGON II Cr. 126, Sta. 36688. 26°00.5'N. 0.88°00.4W,
April 20, 1982, drawn by J. C. Javech; Yarella blackfordi 23.5 mm SL, R/V OREGON II Cr. 126, Sta 36752, 27°30'N, 094°30.3'W, May 9,
1982, drawn by J. C. Javech; Woodsia nonsuchae 1 1.5 mm SL, Eastropac, Sta. 75.225, drawn by J. C. Javech; and Ichthyococcus ovalus 18.1
mm SL, R/V OREGON II Cr. 126, Sta. 36746, 27°59.9'N, 0.88°00'W, May 8, 1982, drawn by J. C. Javech.
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
193
Fig. 100. Lateral views from top to bottom: Bonapar/ia pedaliota 1 1 .5 mm SL. R/V OREGON II Cr. 1 26, Sta. 36688. 26°00.5'N, 088''00.4'W,
April 20, 1982, drawn by J. C. Javech; Margrethia obtusirostra 6.7 mm SL, R/V OREGON II Cr. 126, Sla. 36773. 26°00.rN, 094°00.2'W, May
23, 1982, drawn by J. C. Javech.
were placed in the Infraorder Photichthya. Nearly all have three,
or rarely fewer, proximal pectoral-fin radials, a specialized char-
acter which can be used as a synapomorphy uniting this group.
As noted above, there are a few exceptions which bear four
proximal radials but these appear to be either reversals or are
neomoiphic. Within the Infraorder Photichthya the stomiiform
genera with elongate gill rakers in adults were placed in the
Family Photichthyidae comprised of the genera Polymetme.
Yarella. Pollichthys. Pholichlhys. Vinciguerria. Woodsia, and
Ichthyococcm. This placement was done on the basis of the
presence of Type Gamma photophores in at least most of the
genera, a specialization for the group (as well as for at least some
of the stomiid genera) and therefore a synapomoiTahy. The pres-
ence of elongate gill rakers in this group is not a synapomorphy
because it is primitive for the group.
Essentially, Ahlstrom and Weitzman disagreed on the rela-
tionships of three genera. Alhstrom's Group ( 1 ) was mostly
equivalent to Weitzman's Photichthyidae but included three
genera, Cyclothone, Dtplophos (including Manducus), and Trip-
lophos. placed in the Gonostomatidae by Weitzman. Otherwise,
Weitzman's Gonostomatidae was equivalent to Ahlstrom's
Group (2). Based on evidence available to Ahlstrom and Weitz-
man, on some supplementary evidence provided by Fink and
Weitzman (1982), and on some of our own data, we here present
a somewhat different arrangement based on a more rigorous
phylogenetic analysis than done by Weitzman (1974). It turns
out that Weitzman's analysis of the Stemoptychidae and its
genera is consistently phylogenetic but that of outgroup sto-
miiforms is not. Ahlstrom (1974) did not attempt to analyze
his groups phylogenetically. The evidence available now seems
to resolve the conflict between Ahlstrom (1974) and Weitzman
(1974). However, we would note that the analysis below is to
be regarded as a guide to future studies rather than any sort of
well-corroborated phylogeny. Parts, at least, of the arrangement
need much additional study. Furthermore, the relationships of
the genera in contention by Ahlstrom and Weitzman are still
not fully clear. Some of these genera, Manducus, Diplophos. and
perhaps Triplophos, are relatively primitive within stomiiforms
with few characters specialized beyond the stomiiform level.
This makes placing them in stomiiform subgroups difficult. Cy-
clothone is more derived but retains certain primitive stomi-
iform features and its relationship, although in our view is un-
doubtedly with the gonostomatids, at this time is somewhat
uncertain because our data are not fully analyzed.
The conflict between Ahlstrom (1974) and Weitzman (1974)
arose in part because they both utilized one or the other of
certain characters. Type Beta photophores and "white" pho-
tophore development, as though they were shared specialized
characters, synapomorphies indicating relationships. Instead,
we believe these features are plesiomorphous for stomiiform
subgroups and cannot be used to support a hypothesis of rela-
tionships among stomiiform genera. Our current analysis is as
follows.
Fink and Weitzman (1982:69-75) list and discuss eight syn-
apomorphies for stomiiform fishes. One of these, stomiiform-
type photophores, was described in some detail based in part
on Bassott (1966, 1971). Bassot (1966:574-576), Weitzman
( 1974:338), and Fink and Weitzman ( 1 982:70) recognized Type
Beta photophores as primitive for stomiiforms. Bassot (1966,
1971) recognized two other types of photophores. Type Alpha
194
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Fig. 101. Lateral view from top to bottom: Gonosloma bathyphilum 1 1 .0 mm SL modified after Ahlstrom ( 1 974); Gonostoma elongatum 9.8
mm SL modified after Ahlstrom (1974); Gonostoma ebeUngi 15.0 mm SL modified after Ahlstrom (1974); Gonostoma atlanticum 12.0 mm SL
modified after Ahlstrom (1974); Gonostoma denudalum 20.7 mm SL modified after Sanzo (193 Id).
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
195
Fig. 102. Cydothone signata 9.0 mm SL, drawn by H. Orr.
and Type Gamma, as being more specialized. This recognition,
although not stated by these authors, is based on a concept that
Types Alpha and Gamma photophores of some stomiiformes
appear to be elaborations of Type Beta photophores. In other
words, their particular features appear to be developmental ter-
minal additions to Type Beta photophores and are therefore
available for use as synapomorphies for stomiiform subgroups.
Although more detailed analyses of these features are needed,
for the sake of discussion we here accept that Type Beta pho-
tophores are primitive for stomiiforms.
Weitzman (1974:338), on the basis of outgroup comparison
(not described or discussed in his text), considered four proximal
pectoral-fin radials to be primitive for stomiiforms, their re-
duction to three or fewer to be specialized. We see no reason
to change that analysis. Thus three or fewer proximal pectoral-
fin radials are available as synapomorphous characters for sto-
miiform subgroups.
Ahlstrom ( 1 974:660) described what can be labeled as "white"
photophore development in which most, or at least the ventral
series of photophores, are "laid down initially during a white
photophore stage [before black pigment develops] and only a
few photophores are late forming." One form or another of
"white" photophore development is common to all stomiiforms
except those including the gonostomatid genera Bonapariia.
Margrethia. and Gonostoma, and the stemoptychids of Weitz-
man ( 1 974). Members of these gonostomatid and stemoptychid
genera have a protracted metamorphosis from the larval stage
as well as a gradual, more extended photophore formation. This
latter type of photophore development appears to be an elab-
oration of "white" photophore development and thus we con-
sider white photophore development primitive with respect to
the more complicated forms having prolonged photophore de-
velopment. Again, much information of an anatomical and de-
velopmental nature remains to be gathered from the process of
photophore development.
If "white" photophore development and Type Beta photo-
phores are primitive in regard to stomiiform subgroups and
therefore unavailable as synapomorphies for stomiiform
subgroups, then the conflict regarding the distribution of char-
acters among taxa between Ahlstrom (1974) and Weitzman
(1974) disappears in a phylogenetic analysis by somewhat al-
tering certain of the groups of both authors as follows.
In our tentative scheme of relationships, Weitzman's Ster-
noptychidae and Ahlstrom's Group (2) genera (Ahlstrom, 1 974:
671), Bonapariia, Margrethia, and Gonostoma, the Gonosto-
matidae in the strictest sense, are united by a synapomorphy
consisting of a specialized form of prolonged metamorphosis
and photophore development described by Ahlstrom (1974:
660-661). See also Tables 52-54 herein. These three gonosto-
matid genera and Cydothone apparently share derived char-
acters of the jaws and associated head parts which will be ex-
plained in a later contribution. These four genera retain the
primitive Type Beta photophores, a character relating stomi-
iforms only at the ordinal level. In our opinion these four genera
constitute the Gonostomatidae and Cydothone may have lost
prolonged photophore development through paedomorphic re-
versal associated with the small size of most of its members, a
situation needing further study.
The Stemoptychidae have specialized Type Alpha photo-
phores and the several other synapomorphies listed by Weitz-
man ( 1 974:446-448). In addition they apparently share a unique
photophore growth pattern previously unrecorded. One of us
(Weitzman) has been studying photophore development in re-
lation to phylogenetic studies in stomiiforms and has found that
each cluster or group of photophores of the stemoptychids ap-
pears to develop by budding from one single photophore rather
than by fusion at a later growth stage of separately developed
photophores. This is a terminal developmental addition in pho-
tophore ontogeny and both outgroup comparison and devel-
opmental information indicate that this pattern of photophore
formation is a specialization in comparison to the simpler ap-
pearance of single, separate body photophores (usually one per
scale in any given series found in other stomiiforms). This growth
character appears to be present in all stemoptychid genera for
which we have developmental information. It is therefore a
likely synapomorphy for the group.
Manducus (based on the type species, Gonostoma maderense
Johnson) is a primitive stomiiform, having ordinal-level char-
acters with no known specialized characters except the absence
of an adipose fin and a short neural spine on the preural centmm.
The latter may be a primitive rather than a specialized stomi-
iform feature. Diplophos (based on the type species Diplophos
taenia Gunther) appears to have a transitional stage pectoral
radial morphology between Manducus on the one hand and the
Photichthyidae of Weitzman (1974) (an ill-defined group) and
the Stomiidae on the other. In Manducus the cartilages and
bones of proximal pectoral-fin radials III and IV remain separate
whereas Diplophos has the cartilages, but not the bones, of the
two elements fused. Fink and Weitzman (1982:65-67). In the
"photichthyids" and stomiids the cartilages and bones of the
two medial pectoral-fin radials are fused. This represents the
terminal condition in the transition series except that in some
genera there is a reversal of radial numbers and in Eustomias
there occurs a further specialized, reduced pectoral-fin radial
196
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
AHLSTROM ET AL.: GONOSTOMATIDAE, STERNOPTYCHIDAE
197
Fig. 104. Lateral views from top to bottom: Polyipnus polli 5.2 mm SL R/V GERONIMO Cr. 2, Sta. 155, 05°28S, 01°120'E, August 21,
1963, drawn by J. C. Javech; Argyropelecus hemigy'mnus 7.8 mm SL modified after Sanzo (193 Id); and Sternoptyx sp. 8.8 mm SL. drawn by H.
C. Orr.
Fig. 103. Lateral views from top to bottom: Araiophos eastropas 8.8 mm SL modified after Ahlstrom and Moser (1969); Maurolicus muelleri
10.8 mm SL modified after Ahlstrom (1974); Danaphos oculatus middle metamorphosis modified after Ahlstrom (1974); Valenaennettus tri-
putulutatus middle metamorphosis modified after Ahlstrom (1974); Argyripnus atlanticus 18.7 mm SL modified after Badcock and Merrett
(1972); and maurolicine Alpha 7.5 mm SL modified after Ahlstrom (1974).
198
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
condition. The "photichthyids" and stomiids have specialized
Type Gamma photophores, although it is not known that all
genera in these groups have Type Gamma photophores; this is
a problem for further investigation. Manducus and Diplophos
retain Type Beta photophores and all of these fishes apparently
retain "white" photophore development of one kind or another.
These two characters are only useful at the ordinal level as
synapomorphies. Again, further research on "white" photo-
phore formation is needed since there appears to be more than
one form of this development.
The monotypic Triplophos may or may not be related to
Manducus and/or Diplophos. Triplophos has a variety of derived
features not shared by Manducus or Diplophos. However, this
tells us nothing about its possible relationships with these gen-
era. Triplophos has four proximal pectoral-fin radials but with
some reduction in radial IV, Type Beta photophores, and prob-
ably "white" photophore development, the last two characters
synapomorphous only at the ordinal level. Four pectoral-fin
radials are not a synapomorphy for stomiiforms at any level
since the feature is found in most teleost outgroups. Triplophos
appears to be a primitive stomiiform with certain autapomorph-
ic features associated with an elongate body. Its relationships
are uncertain and there may be indications in the head and
pectoral girdle anatomy of a relationship with certain photich-
thyid genera. The problem needs much study.
Cyclothone retains Type Beta photophores and "white" pho-
tophore development but has its own specialized features such
as only one pectoral-fin radial. It has a modified head and jaws,
which resemble and are, in our opinion, synapomorphous with
those of Gonostoma. The single pectoral-fin radial might be
thought of as a terminal stage in a transition series from Man-
ducus (with four pectoral-fin radials) to Diplophos to some mem-
bers of the "Photichthyidae" and then to Cyclothone. However,
Cyclothone does not have specialized Type Gamma photo-
phores of the "photichthyid" genera. The phylogenetic rela-
tionships of Cyclothone may not be certain as yet, but in many
respects it bears a resemblance to the three gonostomatid genera
and we favor its placement with these genera. See also discussion
above.
Although we have perhaps resolved the differences between
Ahlstrom (1974) and Weitzman (1974), we have not achieved
a useful phylogeny of most stomiiform groups. Rather, we have
attempted to outline certain suggested hypotheses of relation-
ships to be investigated in the future with additional data. Adult
morphological data of the kind used by Weitzman to define and
relate the stemoptychid genera are available in abundance and
may be useful for other stomiiform groups. A closer look at
growth stages with the specific purpose of looking for possible
developmental specializations and terminal additions to char-
acters found in outgroups should greatly aid in delineating re-
lationships among the stomiiform genera. However, problems
associated with a high percentage of homoplasy can be expected
for some groups. The answers to problems of stomiiform in-
terrelationships will not come easily.
Consideration of certain features is in order. For example,
larvae of Diplophos superficially resemble those of Chauliodus
with their prolonged development to a large larval size and great
elongation with bodies that are circular in cross section. Are
these convergent larval specializations or primitive stomiiform
features found only in certain stomiiform genera? The ventral
pigmentation on the body of developing Diplophos resembles
that of developing paralepidids and myctophoids. Is this a prim-
itive stomiiform feature of Diplophos shared with certain sto-
miiform outgroups or a gross convergence of pigment patterns?
Woodsia and Ichthyococcus share with certain stomiid genera
(for example, Eustomias) such developmental features as elon-
gate pectoral-fin rays, trailing guts, pigmentation patterns, and
bodies with a circular cross section. Some, if not all, of these
may be shared larval specializations. But again, independent
appearance of these characters indicated by a high degree of
homoplasy may be a vexing problem. Larvae of other genera
such as Vinciguerria. Pollichthys, and Cyclothone have body
shapes and certain other features that closely, but presumably
superficially, resemble those of clupeoid larvae. Detailed com-
parisons of these similarities may possibly distinguish between
homology and convergence among these taxa.
In summary, a future phylogenetic analysis based on much
additional data may clear up many of the problems of stomi-
iform generic relationships. However, at present we are left with
numerous phylogenetic problems and assignment of certain gen-
era to family-level groups at this time would be misleading. The
above analysis retains Weitzman's Stemoptychidae. It restricts
the Gonostomatidae to the genera Bonapartia. Margrethia, and
Gonostoma. and we recommend the inclusion of Cyclothone.
The other groups of non-stomiid stomiiforms remain unclear
as to family relationships. We agree with Fink and Weitzman
(1982) that Manducus and Diplophos are primitive stomiiforms,
but we cannot provide a stable classification for Manducus.
Diplophos. and Triplophos. Manducus and Diplophos might seem
to be sister taxa because of their similarity of appearance. How-
ever, they share no known specialized character or characters
that would unite them as a stomiiform subgroup except the
absence of an adipose fin and possibly a short neural spine on
the preural centrum. Currently all their other shared characters
seem primitive for stomiiforms. Further analysis of this situa-
tion is needed.
Triplophos is again very much like a primitive stomiiform in
its head especially, but it has a number of specialized stomiiform
features as listed by Grey (1964:106) and may show some re-
lationship to some of the "photichthyid" genera.
That the genera classified in the "Photichthyidae" by Weitz-
man (1974) form some kind of related group seems reasonable.
However, relationships among these genera are not known. That
these "photichthyid" genera are related to Diplophos is possible,
and that the stomiids are related to the "photichthyids" is, in
our view, very probable. The larval specializations of Woodsia
and Ichthyococcus noted above, may be important here because
they may be synapomorphies relating these genera to the sto-
miids.
Until the developmental and adult morphological features of
many stomiiform genera are analyzed in detail, certain aspects
of their developmental stages outlined, and detailed outgroup
analysis performed on all putatively useful characters, we can
make no certain predictions about relationships and classifi-
cation.
(W.J.R.) National Marine Fisheries Service, Southeast
Fisheries Center, 75 Virginia Beach Drive, Miami,
Florida 33149; (S.H.W.) Division of Fishes, National
Museum of Natural History, Smithsonian Institution,
Washington, D.C. 20560.
Giganturidae: Development and Relationships
R. K. Johnson
THE Giganturidae contains two highly-specialized bathy-
pelagic species placed in two monotypic genera; Gigantura
chum Brauer, 1901 and Rosaura mdica (Brauer, 1901). Adults
now placed in Rosaura were formerly recognized as Bathyleptus
Walters, 1961. Morphological specializations of giganturids are
sufficiently divergent and numerous that the group has usually
been accorded subordinal or ordinal status somewhere within
the group now recognized as basal neoteleosts (Stomiiformes +
"Aulopiformes" + "Myctophiformes," see Rosen, 1973; John-
son, 1982; Fink and Weitzman, 1982).
Giganturids are oceanic and deep mesopelagic or bathypelagic
as juveniles and adults. Most hauls successful for juveniles and
adults have been at depths in excess of 500 m (with closing net
captures as deep as 2,000-2,500 m). There is no evidence for
diel vertical migration. G. chum is tropical, R. indica tropical-
subtropical (sensu Johnson, 1982; 185). Giganturids are un-
known from the Southern Ocean, Pacific Subarctic, temperate
North Atlantic (including Mediterranean), and only a single
specimen (G. chum') is known from the eastern tropical Pacific.
Giganturids are relatively large-bodied with adults of Rosaura
achieving more than 220 mm SL, adults of Gigantura more
than 1 70 mm SL. Giganturids are well-known swallowers with
greatly expandable pouchlike stomachs. Most identifiable gut
contents have been fishes, often single large fish ingested whole
(e.g., Regan, 1925). Transformed giganturids are distinguished
from most or all other teleosts by the following combination of
characters; (A) eyes tubular, directed straight forward, in parallel
with main axis of body; (B) gape of mouth extends far behind
eye; teeth fang-like, unbarbed, recurved, depressible; teeth bi-
serial on each jaw, a medial row of enlarged canines and a lateral,
more irregular row of smaller canines; anteriormost canine in
each jaw recurving anteriad; (C) bases of pectoral fins nearly
horizontal, above the gill openings; pectoral fins with a very
high fin-ray count, 37 to 43 in Rosaura. 30-33 in Gigantura;
(D) caudal forked, middle rays of lower lobe lengthened enor-
mously; in one 120.3 mm SL specimen of G. chuni the fila-
mentous extension of the lower caudal lobe adds 243 mm to
the length of the fish; (E) skin loose, scaleless, with a thick layer
of mesenchymal jelly adding substantially to an overall char-
acteristic flabbiness; (F) stomach a thickwalled blind pouch,
giving rise to the intestine ventrally, near midline; intestine
passing laterad and dorsad, to right, continuing along dorsal
contour of stomach until finally turning ventrad behind poste-
rior terminus of stomach and ending at anal papilla; (G) lack
of pelvic fins, dorsal adipose fin, branchiostegal rays, gill rakers;
loss of most of gill arch elements on arches I-III, but with strong,
recurved teeth on 3rd pharyngobranchial (pb) and 4th pb tooth-
plate; loss of numerous other skeletal elements (cf Regan, 1 925;
Walters, 1961, 1964; Rosen, 1973); and (H) considerable con-
solidation of caudal fin skeleton with two presumably com-
pound hypurals (Rosen, 1973).
Development
Eggs of giganturids are unknown. Larvae are known for both
species but only the larva of Rosaura (a single 8.4 mm specimen.
Fig. 105) has been illustrated (Tucker, 1954). For both species
larvae have commonly been taken in the upper 100 m. The
distributional ranges of larvae and adults are coextensive and
there is no evidence for seasonality in reproductive effort (with
only ca 400 known larval specimens, the data are far from
complete). The sexes are separate and according to Clarke and
Wagner (1976) the females may reach twice the size of males,
although available data are sparse. Osteological examination
has been confined to adults except for those elements visible
and described in Tucker's (1954) astonishingly detailed decrip-
tion of the holotype of Rosaura rotunda. Development is direct
but transformation is abrupt with the change from larval to
adult morphology occurring over the approximate size range of
30-40 mm SL in Gigantura and 40-60 mm SL in Rosaura.
Transformation series are now known for both species (only 8
transforming specimens of Gigantura are known, for Rosaura
the count stands at 34) but these results remain unpublished.
The interim account below is thus based on work in progress.
Gross aspect (Fig. 105). — "Rosaura" larvae are short, deep, glo-
bose, translucent and virtually colorless. The forehead is steep,
the eyes small, round and directed laterad. The snout is pointed.
The body is deepest at a vertical through the center of the
opercle. The pectoral insertion is nearly vertical. A dorsal adi-
pose and distinct partly-stalked 5-rayed pelvic fins are present.
Large, readily visible, rather platelike branchiostegal rays are
present. Raptorial jaw teeth are present in the smallest known
larvae (4 mm SL). Teeth on the jaws are biserial with an inner
series of prominent canines and an outer series of shorter more
broadbased teeth on the premaxillaries and dentaries. There are
2-4 recurved smaller fangs on the basihyal. The maxillary is
included in the gape but is edentulous. The abdominal body
wall is nearly transparent and balloonlike, enclosing an expan-
sive gut cavity. The body form remains essentially unchanged
over a period of larval growth extending to ca 30 mm SL (Gi-
gantura) and to ca 35 mm SL {Rosaura). when transformation
begins. Changes during transformation are striking, as described
below. At all stages— larvae, transforming specimens, and ju-
veniles and adults— the species can be distinguished on the basis
of relative depth of the caudal peduncle. The value of this char-
acter varies ontogenetically but the relative peduncle depth is
always greater in Gigantura.
Meristic characters.— Courtis of fin rays do not differ between
larvae and adults except that semi-stalked pelvic fins (5 rayed)
are universally present in larvae and early transforming speci-
mens but are completely lost during transformation. Values for
anal-fin ray counts (8 to 10 in G. chuni. II to 14 in i?. indica)
and pectoral-fin ray counts (30 to 33 in G. chuni. 36 to 42 in
R. indica) separate the two species without overlap. Dorsal-fin
ray counts ( 1 6 to 19) have the same range in both species. The
caudal is the first fin to form; it is asymmetric with 10 -t- 6(7)
principle caudal rays and (3)4(5) procurrent caudal rays above
and below. Next to form, in order, are the dorsal + anal fins,
pelvic fins, and pectoral fins (the dorsalmost pectoral rays begin
199
200
ONTOGENY AND SYSTEMATICS OF RSHES-AHLSTROM SYMPOSIUM
."//,/.,
Fig. 105. Giganturidae. (Upper) Larva oi Rosaura indica. 8.4 mm SL (=holotype oi Rosaura rotunda from Tucker, 1954). (Lower) Adult
Rosaura indica. 182 mm SL (from Berry and Perkms, 1966).
to differentiate in larvae as small as 5.5 mm SL. but the ventral-
most pectoral rays are the last fin rays to be formed). The pelvic
fins appear just below the dorsal-fin origin and do not greatly
shift in relative position until transformation. A dorsal finfold
connects the incipient dorsal fin with the caudal fin in small
larvae, but loses this connection in larvae larger than 6 mm SL.
and shrinks in extent but remains as a highly visible adipose fin
until transformation, when it is resorbed.
Peritoneal pigment sections. — A single peritoneal pigment sec-
tion characterizes the larvae of both species. This section lies
just above and posterior to the dorsal transverse limb of the
intestine. The section is never paired as in synodontoids and
remains proportionately constant in size throughout larval life
and is represented in adults as a small, intensely-black oval
pigment patch above the stomach (growth of the section ap-
parently ceases at about the onset of transformation, but the
section apparently remains in both juveniles and adults of both
species). The dense brown or black pigment enclosing the gut
is not derived from this peritoneal pigment section, as is true
for many "inioms" (see Johnson, 1 982) but develops separately
dunng transformation (as in Aleptsaurus and Omosudis, Was-
sersug and Johnson. 1976).
Other pigmentation. — \n both species pigmentation in larvae
occurs in three areas (other than the peritoneal section), the
eyes, over the optic lobes, and on the sides of the body posterior
to the dorsal-fin base. In some but not all pre-transformation
specimens of Gigantura, very small punctate melanophores ap-
pear over the still otherwise essentially transparent lateral ab-
dominal body wall.
Gut morphology.— The stomach is enlarged and sac-like. The
mtestine leaves the pyloric region of the stomach, descends
round the left margin of the abdominal cavity, crosses trans-
versely upon the ventral body wall, reascends the right side and
then turns again, descending abruptly and obliquely down and
posteriad to the vent.
Transformation —Changes during transformation are numer-
ous and striking: (A) Body shape. The body changes in shape
from short, rotund and deep, rather as in some ceratioid larvae
JOHNSON: GIGANTURIDAE
201
(Bertelsen, 1 95 1 ) or the larvae of certain scopelarchids (Johnson,
1974b, 1982) to the elongate, shallow, slender shape of the gi-
ganturids. The head while still massive is proportionately much
less so ('/« vs 'A SL in Rosarua) and the dorsal head profile is
essentially horizontal rather than steeply oblique (Fig. 105). (B)
Eyes. Eyes in larvae are round, small and directed laterad; eyes
in adults are fully tubular and directed rostrad. (C) Fins. Dis-
tinct, partly-stalked, 5-rayed pelvic fins are present in larvae,
resorbed or shed during transformation, and lacking in adults.
The line of insertion of the pectoral-fin rays is obliquely vertical
in larvae, essentially horizontal in adults. In larvae the pectoral
insertion is behind the gill slit, in adults (especially prominent
in Gigantura) the pectoral insertion is substantially above the
gill slit. A distinct dorsal adipose fin is present in larvae, absent
in adults. Procurrent caudal fin rays number (3)4(5) in larvae
and are prominent, in adults procurrent caudal rays are fre-
quently embedded in the skin, difficult to see, and number
(0)1(2,3). (D) Teeth. Among the most striking changes occurring
dunng transformation is the total loss of all larval teeth (in-
cluding basihyal teeth). Transforming specimens are character-
ized by a scalloped, irregularly-emarginate jaw edge (upper and
lower) which is edentulous. None of the 40 known transforming
specimens shows development of adult teeth and the smallest
known post-transformation specimen (36.4 mm SL, G. chuni;
47.9 mm SL, Rosaura indica) possess a full complement of
adult teeth. (E) Color. Larvae are essentially translucent with
very little development of pigment, adults are entirely blackish/
brown (often with the development of an iridescent finish in
Gigantura). Onset of transformation is indicated by the "sud-
den" widespread development of pigmentation. (F) Loss of skel-
etal elements. Larvae possess at least the following skeletal ele-
ments not seen in adults: symplectic, coracoid, cleithrum,
posttemporal, supracleithrum, branchiostegal rays.
Relationships
The first association of "Rosaura" with the giganturids was
by Ahlstrom and Berry about 1960 (letters and mss material
made available by H. G. Moser) with the first published sug-
gestion made in Berry and Perkins (1966). Key characters sug-
gesting relationship included the very high pectoral-fin ray count
and the highly unusual 10-1-6(7) distribution of principle caudal
rays, apparently unique to "Rosaura" and the giganturids. The
disparities between "Rosaura" larvae and adult giganturids—
briefly outlined above— left doubt in many minds, but the cap-
ture of essentially complete transformation series (to be de-
scribed and illustrated in detail elsewhere) make it unquestion-
able that "Rosaura" is the larval form of the giganturids. With
a caudal peduncle depth of ca 9.9% of SL (Tucker, 1954:168)
there is likewise no doubt that the type of Rosaura rotunda
represents a larva of "Balhyleptus," requiring recognition of the
more elongate, shallow-bodied species as Rosaura indica (Brauer,
1901). The deeper-bodied species is Gigantura chuni Brauer,
1901 (other species have been described but the characters used
to distinguish them do not work, nor has other evidence been
found to support the hypothesis of more than two species). Of
the two, Walters (1961, 1 964) argued for the more apomorphous
condition of Gigantura but his characters need to be re-exam-
ined in light of outgroup comparisons and in conjunction with
other characters.
Vanous authors have allied giganturids with such disparate
groups as Stylephoridae, Saccopharyngiformes and "... a line
[leading] from a subiniomous group such as the esocoids toward
the synodontoid inioms, and this line later may have given rise
to the Cetunculi . . ." (Walters, 1961). Rosen (1973:438-441)
has offered evidence that the original placement by Regan (1925:
57) of giganturids with synodontoids was correct. Rosen calls
particular attention to similarities in upper jaw and infraorbital
configuration with synodontoids and the presence of a retractor
dorsalis (=RAB in Rosen, 1973; see Winterbottom, 1 974b) mus-
cle configration state characteristic of the synodontoid/alepi-
sauroid line (Johnson, 1982:85, 95). An important character
(Johnson, 1982:71; Okiyama, this volume) uniting synodon-
toids with alepisauroids is the presence in larvae of multiple (3
or more) peritoneal pigment sections. Uniting synodontids and
harpadontids (sensu Sulak, 1977) is the fact that in larvae of
these fishes the sections are paired . . . and not connected over
the gut. The condition in "Rosaura" is that seen in aulopids,
chlorophthalmids, primitive scopelarchids, and ipnopids, viz. a
single section situated over the gut. This is the state thought
primitive for inioms. Also distinguishing the giganturids is a
unique conformation of the gut. In larvae the gut arises from
the pylorus, descends round the left margin of the abdominal
cavity, crosses transversely midventrally, reascends the right
side, turns abruptly mediad, then turns again, descending abruptly
and obliquely to the vent. In adults the intestine arises mid-
ventrally, makes a few small twists, ascends the right side, and
passes posteriad above the dorsal contour of the expanded stom-
ach, only descending to the vent posterior to the terminus of
the stomach. In all the inioms I have examined the intestine
arises midventrally and passes essentially straight back to the
vent along the midventral wall of the abdominal cavity. For the
time being, the available evidence suggests that the giganturids
are neoteleosts (retractor dorsalis muscle), allied with the inioms
(discrete peritoneal pigment section), diverging early from the
rest and acquiring characters making them among the most
specialized and distinctive of teleosts.
Field Museum of Natural History, Roosevelt Road at
Lake Shore Drive, Chicago, Illinois 60605.
Basal Euteleosts: Relationships
W. L. Fink
AS mentioned in the introduction to this section of the sym-
posium, the order Salmoniformes has had a history of
attrition, such that today I would recognize it as coextensive
with the Salmonidae. Previously included taxa are now scat-
tered, primarily as unresolved lineages at or near the base of
the Euteleostei. What follows is a preliminary analysis, a sketch
of alternative hypotheses of interrelationships of the basal eu-
teleosts. Fully resolving these problems will take more time and
more material than I have had available to me, and I hope that
work stimulated by this symposium will provide insights which
have not been forthcoming using traditional material and char-
acters.
Unfortunately, very little information of a comparative nature
is available on the larvae of basal euteleosts, and when these
larvae have been discussed, only rarely have characters or char-
acter transformations useful at large clade levels been men-
tioned. Since adult specimens are more easily available in most
collections, that is what I have relied on, with examination of
larvae when possible.
Results
The Euteleostei is a large group of modem teleosts which is
poorly diagnosed in terms of unique traits, and most more phy-
logenetically advanced members lack some of the diagnostic
characters. Patterson and Rosen (1977) considered the following
as euteleostean traits: 1) an adipose fin, 2) nuptial tubercles, and
3) an anterior membraneous component to the first uroneural.
Near the "base" of the Euteleostei, Fink and Weitzman(1982)
recognized several lineages, including the Esocoidei, Ostario-
physi, Argentinoidei, Osmeroidei, Salmonidae, and Neoteleos-
tei. All were considered monophyletic, but the interrelations of
these large clades were left unresolved (Fig. 106). Below is a
review of each of the groups, with new information included
when possible.
Esocoidei or Esocae.—The^ fishes have been a continuing
problem for ichthyologists. They are considered as euteleosts
on the basis of an anterior membraneous component to the first
uroneural, although it is not extensive. No esocoids can have
an adipose fin as the dorsal fin is posteriorly situated. Neither
do they have breeding tubercles. Rosen (1974) provided diag-
nostic characters documenting monophyly of the group. Fink
and Weitzman (1982) suggested that esocoids could be the sister
group of all other euteleosts based on the lack in the latter of a
toothplate on the 4th basibranchial, a bone which is present in
esocoids and other primitive teleosts (see those authors for a
discussion of the distribution of this character). Wilson and
Veilleux ( 1 982) have recently reviewed interrelationships in the
Umbridae, and they place Umbra and Dallia as sister taxa, with
Novumbra as their sister group; all these together are placed as
the sister group of Esox. This corroborates the hypothesis of
Nelson (1972).
Rosen (1974) considered Lepidogalaxias to be a member of
this assemblage, which he termed the Esocae. Fink and Weitz-
man (1982) questioned that hypothesis, leaving the genus un-
placed. I have further comments and a new hypothesis of its
relationships below. I have nothing to add to what Fink and
Weitzman (1982) did with esocoids sensu stricto. and until more
is forthcoming, consider them the likely sister group to other
euteleosts.
Ostariophysi— In terms of numbers of species and morpholog-
ical diversity, this is the dominant basal euteleostean group.
Fink and Weitzman (1982) did not consider the relations of
these fishes to other euteleosts, primarily because their survey
was intended to establish the placement of stomiiforms, and
there was no evidence suggesting relationship between the two
groups. No phylogenetic examination of ostariophysan rela-
tionships to other teleosts has been done since Rosen and Green-
wood (1970) expanded traditional concepts of the group by
adding the previously protacanthopterygian gonorynchifonns.
Fink and Fink ( 1 98 1 ) examined relationships within the group,
placing siluroids and gymnotoids as sister taxa (order Siluri-
formes), these the sister taxon of characiforms, and these to-
gether the sister group of cypriniforms (the Otophysi, inclusive);
sister group relationship of the gonorynchiforms to the Otophysi
was corroborated. This entire assemblage was considered mono-
phyletic on the basis of numerous characters, including lack of
a dermopalatine, unique gasbladder morphology, specializa-
tions of the vertebrae, and adductor mandibulae anatomy.
Argentinoidei.— Cvetn'wood and Rosen (1971) combined the
alepocephaloid and argentinoid fishes into an expanded Argen-
tinoidei, in the Salmoniformes. Fink and Weitzman (1982) agreed
with the combination of the two groups and used the formal
subordinal name to include both subgroups. However, Fink and
Weitzman (1982) were unable to provide evidence bearing on
relationships of these fishes, even though their cladogram (Fig.
23, Fig. 106 herein) showed them as the sister group of the
osmeroids. I have similarly been unable to place them, in part
because of lack of adequate material.
Osmeroidei— Thii group, which includes the northern and
southern smelts, galaxiids (here including Lovettia and Aplo-
chiton), Plecoglossus, and salangids, can be diagnosed as mono-
phyletic based on several characters, including presence of one
or more rows of teeth near the medial border of the mesopter-
ygoid, loss or appearance late in ontogeny of the articular bone,
and presence of a foramen in the posterior plate of the pelvic
bone. Some subgroups of osmeroids have lost various of these
diagnostic characters, but the patterns of loss allow other fea-
tures to provide evidence of relationship in the group.
Nevertheless, relationships within the suborder remain prob-
lematical. The following review is based upon examination of
specimens, the literature, and the contributions to this sym-
posium. Incidentally. I have not attempted to diagnose the var-
ious genera, but McDowall's comments (this volume and 1 969)
indicate that such needs to be done. The phylogenetic hypoth-
202
FINK: BASAL EUTELEOSTS
203
ESOCAE
OSTARIOPHYSI
ARGENTINOIDEI
OSMEROIDEI
NEOTELEOSTEI
SALMONIDAE
STOMIIFORMES
AULOPIFORMES
MYCTOPHIFORMES
ACANTHOMORPHA
Fig. 106. The hypothesis of relationships suggested by Fink and
Weitzman ( 1 982) for the basal euteleosts.
eses and data are included in Fig. 107 and its caption. The data
used in this analysis were chosen partly because they have been
used traditionally in osmeroid systematics but I have little con-
fidence in some of them; as a result this analysis represents a
preliminary sketch of a more detailed study.
The most striking thing about osmeroid systematics is that
we still have questions about some very basic things, such as
the status of the Osmeridae. As noted by Nelson (1970) and
Rosen (1974), no evidence has ever been presented that the
family is a monophyletic group. Indeed, it seems quite possible
that Plecoglossus could be more closely related to some "os-
merids" than to others, and this would render the family para-
phyletic. A minimal requirement of any future work on system-
atics of the group should be documentation of whether it is
natural.
Fig. 107. Alternate cladograms of relationships within the Osme-
roidei. The bottom figure represents the hypothesis supported when all
characters are given equal weight and paedomorphic traits are consid-
ered homologous. The top figure represents the hypothesis which con-
siders the paedomorphic reductive traits of salangids and galaxiids as
non-homologous. For discussion, see text.
The supporting characters are listed below, with the derived condition
indicated by a 1 , the primitive by a 0. Each character number is indicated
on the cladogram where it is in the derived state. Dark squares indicate
unique appearance of a trait; empty squares indicate multiple appear-
18-20
6ALAXIIDAE
lovettia
Aplochiton
Retropinna
Stokellia
Prototroctes
SALANGIDAE
Plecoglossus
"OSMERIDAE"
2>29 -J-.GALAXIIDAE
lovettia
'Aplochiton
■SALANGIDAE
-Retropinna
•Stokellia
Prototroctes
Plecoglossus
'OSMERIDAE"
ance of a trait; triangles indicate a trait that is reversed at a lower level
of generality; and circles indicate those characters in the reversed state.
1 . Posterior shaft of vomer (0) long ( 1 ) shori. 2. Articular bone (0)
present and fused with angular (1) absent or greatly reduced. 3. Meso-
pterygoid teeth (0) over much of bone ventral surface (1) restricted to
medial border of ventral surface or lacking. 4. Pelvic foramen (0) absent
( 1 ) present. 5. Anchor membrane of egg (0) absent ( 1 ) present. 6. Caudal
skeleton fusion patterns (0) none or rudimentary neural arches fusing
with centrum and then, if at all, to the uroneural ( 1 ) rudimentary neural
arch fusing with uroneural first, then these to the centrum. 7. Infraorbital
sensory canals (0) curved posterodorsally ( 1 ) curved posteroventrally.
8. Mesocoracoid (0) present ( 1 ) absent. 9. Dorsal fin position (0) forward
(1) posterior. 10. Principal caudal fin rays (0) 10/9(1) 9/9 or fewer. 1 1.
Palatine teeth (0) present (1) absent. 12. Ectopterygoid bone (0) present
( 1 ) absent. 1 3. Extrascapular (0) present ( I ) absent. 14. Coracoid-cleith-
rum process (0) present (1) absent. 15. Posterior pubic symphysis (0)
present ( 1 ) absent. 1 6. Scales (0) present ( 1 ) absent. 1 7. Vomerine teeth
(0) present (1) absent. 18. Posterior border of bones of suspensorium
(0) smooth (1) deeply incised or emarginate. 19. Principal caudal fin
rays (0) 9/9(1) 8/8. 20. Hypural number (0) 6 (1) 5. 21. Infraorbital
sensory canals (0) not extending to preopercle (1) extending to pre-
opercle. 22. Ceratohyal ventral border (0) more or less straight, bran-
chiostegals along most of its length ( 1 ) deeply concave anterioriy, bran-
chiostegals restricted to area posterior to concavity. 23. Homy abdominal
keel (0) not present ( 1 ) present. 24. Ovaries (0) both present ( 1 ) left only.
25. Ectopterygoid bone (0) posterior to autopalatine (1) ventral to au-
topalatine (coded as present in Stokellia based on McDowall, 1969).
26. Cucumber odor (0) absent (1) present. 27. Basioccipital lateral pegs
(0) none (1) present. 28. Lateral hyomandibular spur (0) not present (1)
present. 29. Caudal fin posterior border (0) deeply forked (1) rounded
or emarginate. 30. Adipose fin (0) present ( 1 ) absent. 3 1 . Mesopterygoid
teeth (see also Character 3) (0) restricted to ventromedial area of bone
( 1 ) absent.
204
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Salangids have been associated in the past with various mem-
bers of the osmeroid assemblage, but even this was questioned
by Nelson (1970). Rosen (1974) presented evidence from the
caudal skeleton which shows that salangids are osmeroids, but
no evidence about their placement within the group has been
presented to date. Fink and Weitzman (1982) agreed with Rosen
and placed the Salangidae as incertae sedis in the Osmeroidei.
What little evidence I have been able to find about the rela-
tionships of salangids is equivocal. If examined by a standard
parsimony procedure, as represented by the Wagner analysis
shown in Fig. 107 (bottom), the numerous reductive traits of
salangids place them within the "southern smelt" plus galaxiid
assemblage. On the other hand, salangids share with Plecoglos-
siis and the "osmerids" a complex caudal skeleton character
involving fusion of uroneural 1 to a compound centrum made
upofPUl, Ul,and U2, followed ontogenetically in some forms
by fusion of rudimentary neural arches with the uroneural por-
tion of the complex. This latter character is in contrast to the
autogenous uroneurals of most galaxiids, the "southern smelts,"
and other primitive teleosts. Further, when uroneurals and ru-
dimentary neural arches are fused in galaxiids, the fusion se-
quence is rudimentary neural arch to the compound centrum,
followed by fusion with the uroneural, rather than the reverse.
The hypothesis that emerges from these observations is illus-
trated in Fig. 107 (top), showing salangids, Plecoglossus. and
"osmerids" in an unresolved trichotomy. For further discussion
of caudal fin morphology, see Greenwood and Rosen (1971),
Rosen (1974), and Fink and Weitzman (1982).
Any choice of these alternate hypotheses of salangid relation-
ships would rest on whether or not one wished to accept the
numerous reductive traits that unite the salangids with the
"southern smelts" and galaxiids as homologues. Such choice is
based on criteria which cannot be discussed in detail at this
point due to space restrictions, but I have commented elsewhere
(Fink, 1982) on hypothesis choice forced by confrontation with
apparent paedomorphosis. In this case, for example, some of
the general morphological attributes that salangids share with
the members of those groups differ when examined in detail.
Although this lack of close correspondence in similarity is cer-
tainly no guarantee that the reductions are not homologous, it
does raise the issue. Further, the highly developed caudal skel-
eton of salangids is identical to that of "osmerids," and thus
more differentiated than that of either the southern smelts or
galaxiids. This incongruity in degree of morphological differ-
entiation suggests that in this case, one should be cautious in
assuming homology in the reductive process and search for
other, non-reductive characters to resolve possible misplace-
ments.
The family Sundasalangidae is not accepted herein because
in every case in which Roberts (1981) contrasted sundasalangids
and salangids, the character for salangids was primitive. I suggest
that recognition of family rank for Sundasalanx^^oviXd probably
render the Salangidae paraphyletic and thus defined only by the
absence of characters present in Sundasalanx. This is unac-
ceptable both because it forces recognition of a group based on
characters its members lack and because it artificially breaks up
a group all of whose members share a unique evolutionary his-
tory.
Regarding the "southern smelt assemblage" (including gal-
axiids, but excluding salangids), I am less pessimistic than
McDowall (this volume). I have taken the liberty of using the
data he has presented and combined them with my own limited
survey of specimens and the literature to produce the hypotheses
shown in Fig. 107. The group can be diagnosed by presence of
a posteroventral deflection of the infraorbital sensory canal (Nel-
son, 1972) and 9/9 or fewer principal caudal-fin rays (vs a pos-
terodorsal curvature of the canal and 10/9 rays in outgroups).
Several characters support the placement of Retropinna and
Prototrocles as sister taxa including presence of an abdominal
homy keel, loss of the right ovary, and ceratohyal morphology.
I have no specimens o( Stokellia on hand, but McDowall's work
( 1979) clearly shows that the genus is diagnosable and that it is
related to Retropinna and Prototrocles. Unfortunately, when
contrasted with Stokellia, it is not clear that Retropinna is di-
agnosable, since the latter is then differentiated by primitive
characters present in other taxa.
Relationship among Aplochiton, Lovettia and the galaxiids is
supported by numerous characters, as shown in Fig. 107. I have
been unable to find any features that link the former two genera
together, however, and more work needs to be done with them.
Galaxiids themselves can be shown to be monophyletic based
on such characters as basioccipital "pegs" extending lateral to
the anterior centrum (McDowall, 1969, Figs. 2B, lOA, but note
lack of "pegs" in G. paucispondylus. Fig. 1 OB).
In summary, it is suggested that the broad outlines of rela-
tionships among the osmeroids are beginning to emerge, much
as suggested by Gosline (1960a), with a "southern smelt" as-
semblage and an "osmerid" assemblage. Interrelationships within
these groups remain problematical, the most obvious problems
being establishment of the natural groups within the "osmerids"
and placement of the salangids.
Salmontds. — M.onox>\\y\y of this group is based primarily on a
single character, apparent polyploidy of the karyotype (Gold,
1979). Several investigators have studied interrelationships of
salmonids, most notably Behnke (1968) and Norden (1961), but
these works were not phylogenetic and changes can be expected.
I have examined phylogeny within the group only to establish
polarities for characters relevant to relationships with other te-
leosts. Regarding the latter relationships, there have been several
opinions, with most workers approaching salmonids with an
eye to finding ancestors ofother groups (see, e.g., Gosline, 1960,
Diagram 2). The only phylogenetic analysis to date is that of
Rosen (1974), which was discussed by Fink and Weitzman
( 1 982). The latter authors presented data which they considered
suggestive of neoteleostean relationship for salmonids: presence
in some members of paired cartilages anterior to the ethmoid
region (resembling the median rostral cartilage of neoteleosts)
and the exoccipital forming part of the occipital condyle. The
anterior cartilages were reported by Fink and Weitzman (1982)
to be prominent in Prosoplum, an observation which I can
confirm from additional specimens. In addition, examination
ofsmall juvenile cichlids shows that the rostral cartilage appears
to develop ontogenetically from bilateral cartilage bodies which
fuse at the midline; this is suggestive of corroboration of Fink
and Weitzman's (1982) hypothesis that the rostral cartilage
evolved from paired cartilages anterior to the ethmoid region
like those in Prosoplum. More work needs to be done on the
homology of "accessory" ethmoid cartilages, using double stain-
ing techniques and histology on a wide variety of teleosts.
1 can also add to what Fink and Weitzman ( 1 982) noted about
the occipital condyle. 1 have confirmed that the exoccipital forms
part of the condyle in Thymallus and "salmonins." This mor-
phology is also present in Prosoplum. but is lacking in other
FINK: BASAL EUTELEOSTS
205
coregonins. In a number of features, including the morphology
of the nares, Prosopium stands as the sister group of other cor-
egonins, and this, plus the presence in the outgroup Salmoninae
and Thymallus of exoccipital participation in the condyle, im-
plies that phylogenetically derived coregonins have secondarily
lost that morphology. As noted by Fink and Weitzman (1982),
the condyle structure as found in salmonids is found also in
neoteleosts. It is also present in Lepidogalaxias (see below) and
in some osteoglossomorphs. I do not wish to belabor the possible
importance of this character, especially since more careful on-
togenetic and morphological studies need to be done and other
characteristics evaluated.
A few observations from my survey of salmonids may be
added here. I have found but two characters in the literature
which diagnose the coregonins; one of these needs modification
and the other needs to be more concisely put. Lack of maxillary
teeth has been used to diagnose the group, relative to other
salmonids (Norden, 1 96 1 ), but this needs to be emended to lack
of the teeth in adults, since I have found maxillary teeth in
Prosopium of around 19 mm SL. I have not yet examined spec-
imens this small of other coregonins so do not know the gen-
erality of this primitive state. The other character is reduction
in the teeth in general; this needs to be quantified relative to
the outgroups.
The salmonins and Thymallus can be placed together based
on lack of ossification of the supraethmoid (hypethmoid of Nor-
den, 1961; Behnke, 1968), and apparently on yolk character-
istics, and larval size (Kendall and Behnke, this volume). Re-
garding other relationships within salmonids, I have nothing to
add.
Lepidogalaxias.— The position of Lepidogalaxias is controver-
sial. I remain unconvinced by Rosen's (1974) hypothesis that
the genus belongs with the esocoids. When I previously dis-
cussed this genus (Fink and Weitzman, 1982), I had not seen
any specimens, but R. M. McDowall has generously made sev-
eral available for dissection and clearing and staining. There is
no question that this little fish is a potpourii of contradictory
and reductive characters and it is no wonder that it has been so
difficult to place. Pursuing the potential of relationship of this
species to galaxiids, extensive comparisons with members of
that group have been made. Lepidogalaxias shares a host of
reductive characters with galaxiids. NVhile these may indeed be
synapomorphous traits, in cases where extensive paedomor-
phosis is suspected, and this appears to be so in the morpho-
logical similarities involved, one hopes to find some innovative,
non-reductive characters which supply evidence for grouping.
I have found two such characters which suggest that Lepido-
galaxias is related to neither esocoids nor osmeroids, but rather
may be the sister group of the Neoteleostei, as diagnosed by
Rosen (1973) and Fink and Weitzman ( 1 982). This is supported
by the presence in Lepidogalaxias of two non-reductive traits,
a retractor dorsalis muscle and occipital condyle composed of
both the basioccipital and exoccipital bones. As discussed just
above and by Fink and Weitzman ( 1 982), the latter trait is also
shared with salmonids. Lepidogalaxias lacks a rostral cartilage
or its homologue and type 4 teeth (hinged teeth with a posterior
axis of rotation. Fink, 1981) and this would prevent its place-
ment within the neoteleostean assemblage. Placing Lepidoga-
laxias as the neoteleostean sister group and leaving salmonids
as their sister taxon presumes either that rostral cartilage homo-
logues in the salmonids have been lost in Lepidogalaxias or are
EUTELEOSTEI
NEOTELEOSTEI
EURYPTERYGII
OSTEOGLOSSOMORPHA
ELOPOMORPHA
CLUPEOMORPHA
ESOCOIDEI
OSTARIOPHYSI
ARGENTINOIDEI
OSMEROIDEI
SALMONIDAE
LEPIDOGALAXIAS
STOMIIFORMES
AULOPIFORMES
MYCTOPHIFORMES
ACANTHOMORPHA
Fig. 108. Summary cladogram of relationships and characters dis-
cussed in the text.
not homologues after all. This ambiguity is reflected in Fig. 108
by a trichotomy. Clearly, more work remains to be done before
we can be really confident in the phylogenetic placement of this
intriguing fish.
Lepidogalaxias can be diagnosed by a number of characters,
the most striking of which is fusion of the frontal bones into a
single ossification (Rosen, 1974, Fig. 40B). In their comments
on this species. Fink and Weitzman ( 1 982) noted that there was
a disagreement about whether there are mesopterygoid teeth
present; Rosen's statement that teeth are lacking is cortect.
Stomiiformes. — Vmk and Weitzman (1982) recently examined
the monophyly and relationships of stomiiforms to the other
basal euteleosts and corroborated Rosen's (1973) hypothesis
that they are the sister group to the rest of the Neoteleostei,
removing them from the "salmoniforms." This placement is
supported by several apomorphic traits, including presence of
retractor dorsalis muscles and type 4 tooth attachment, as well
as exoccipital participation in the cranial condyle and a rostral
cartilage. Weitzman (1974) presented a hypothesis of relation-
ships at the "family" level within the stomiiforms, as well as a
detailed phylogeny of the Stemoptychidae. In this volume, I
present a generic-level phylogeny for the barbeled stomiiforms
(Family Stomiidae) and some brief comments on the "gonosto-
matid-photichthyid" genera. Weitzman is currently working on
relationships of the latter fishes and has made considerable com-
ments in this volume (see Ahlstrom, Richards and Weitzman,
this volume).
206
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Eurypterygii. — FinaWy, a few comments are due on the Myc-
tophoidei of Greenwood et al. (1966). This group was disman-
tled by Rosen (1973), and divided into two large groups, Au-
lopiformes and Myctophiformes. These two groups, together
with the Paracanlhopterygii and Acanthopterygii, were classified
into a new group, Eurypterygii. Aulopiformes was placed as the
sister group of all other eurypterygians, and myctophiforms as
the sister group to paracanthoptergyians and acanthopterygians.
All of these, together with stomiiforms, form the Neoteleostei.
Fink and Weitzman (1982) tentatively accepted monophyly of
the Eurypterygii based on the presence in its members of a
toothplate fused with the third epibranchial. Aulopiformes con-
tains a large number of families, including the Giganturidae,
covered in this portion of the symposium. About the latter
family I have little to say except that my own dissections cor-
roborate Rosen's placement of it.
Summary
A summary of the hypotheses I have discussed above is given
in Fig. 108. The most striking aspect of it is the degree of un-
certainty about relationships among the clades. This may be in
part due to the limitations of my study, but it does seem to me
to be a fair summary of the status of well corroborated hypoth-
eses we now have about this level of teleostean phylogeny. There
are certainly other arrangements that can be made, depending
on which characters one wishes to stress, and none of these
should be discarded out of hand. As examples, I will cite two
characters and their implications.
First, lack of the posterior shaft of the vomer suggests that
salmonids and osmeroids are sister taxa. Appropriate outgroups
have the shaft ranging from "moderate" (e.g., Chanos) to "elon-
gate" (argentinoids). My own opinion, based on occipital con-
dyle structure of salmonids, is that the reduction in vomer length
has occurred independently in the two lineages (it has also been
reversed within both); the ultimate value of the occipital char-
acter remains to be seen.
The second character, presence of breeding tubercles, is now
considered a euteleostean trait. Note, however, that tubercles
are lacking in esocoids and argentinoids but are present in os-
tariophysans, osmeroids, and salmonids, indicating that these
three clades form a monophyletic group. Again, there are char-
acters that contradict this grouping, but it nevertheless is worthy
of consideration.
It is always frustrating when one sets out to solve a particular
problem and then comes to the end of the allotted time without
a resolution. Although I have been able to shed some light on
several problems relevant to the goals of this part of the sym-
posium, I have not been able to unravel the interrelationships
among the major basal euteleostean clades. Clearly more work
is needed, especially with character suites which have been tra-
ditionally neglected. Almost all of our concepts of relationships
at this level are based on features of the adult caudal skeleton
and branchial basket. Some work on soft anatomy, particularly
the muscles of the head, has been informative at these levels
and one hopes that other parts of the soft anatomy will be equally
profitable. One area virtually untouched is larval anatomy. It
might be expected that not many important features will be
found because of the preponderance of primitive characters in
larvae. But larval characters have proven useful, as is shown by
the ontogenetic transformation in tooth types in stomiiforms
(from type 4 to type 3; see Fink, 1981) as well as the specialized
fin traits discussed by Ahlstrom et al. (this volume) for argen-
tinoids. It is in both these areas, ontogenetic character trans-
formations and presence of specializations for larval life, that
study of larval fishes promises rewards. The inclusion of larval
morphology in studies of higher level relationships should pro-
vide a richer data base than we currently have and perhaps will
reveal some crucial characters for resolving the basic questions
I have addressed above. This symposium has already stimulated
in a major way the examination of larvae for phylogenetic anal-
yses, and I predict that it, combined with the new ways now
emerging of analyzing ontogenetic information, will mark a new
phase in the modem study of fish classification.
Museum of Zoology, University of Michigan, Ann Arbor,
Michigan 48109.
Myctophiformes: Development
M. Okiyama
MYCTOPHIFORMES is currently adopted as a distinct
order with intermediate affinity between the lower and
higher teleost groups, whereas no one feature would satisfac-
torily separate all of them from all Salmoniformes (Gosline et
al., 1966). Except Rosen (1973), recent workers agree well with
the familial composition of this order despite slight differences
in the familial or subordinal definition.
Table 56 shows the recent classification given by Johnson
(1982) based on the most comprehensive knowledge now avail-
able. Important points of this scheme are the exclusion of Sco-
pelarchidae from Alepisauroidei and Pseudotrichonotidae from
Myctophiformes. Further details in this connection will be men-
tioned again in my paper on relationships (this volume).
Exploitation of the vast hydrosphere covering the pelagic as
well as benthic habitat between the surface and abyssal or ul-
traabyssal plain by diversified members of this group is doubt-
lessly the important aspect in discussing the ontogenetic prob-
lems of the myctophiform lineage. Of the five suborders,
Myctophoidei and Alepisauroidei are exclusively pelagic and
the remaining are demersal including secondary pelagic genera
such as Parasudis and Harpadon. Synchronous hermaphrodit-
ism is common to the deep-water and offshore forms belonging
to Chlorophthalmoidei and Alepisauroidei with the single ex-
ception of Bathysauridae in Synodontoidei (Table 56).
In general, the systematics of this order are rather well under-
stood except for several families or genera. As is clearly shown
OKIYAMA: MYCTOPHIFORMES
207
Table 56. Systematic Status and the Current Knowledge on Early Life Stages in Myctophiformes.
Suborder and family
No.
Reproduc- _
tion''
Information
species
Eggs Larvae
Main sources
7 +
G
_(.C +_|.d
Okiyama (1974b)
Aulopoidei
Aulopidae
Myctophoidei
Myctophidae"
Neoscopelidae
Chlorophthalmoidei
Chlorophthalmidae
Ipnopidae
Notosudidae
Scopelarchidae"
Synodontoidei
Balhysauridae
Harpadontidae
Synodontidae
Alepisauroidei
Alcpisauridae
Anotopteridae
Evermannellidae"
Omosudidae
Paralepididae
Aulopus
Diaphus. etc.
Neoscopelus
Scopelengys
Solivomer
Chlorophthalmus
Parasudis
Bathysauropsis
Ipnops
Bathytyphlops
Bathymicrops
Bathypterois
Ahliesaurus
Scopelosaurus
Luciosudis
Scopelarchus, etc.
Bathysaurus
Harpadon
Saurida
Synodus
Trachinocephalus
Alepisaurus
Anotopterus
Evermannella. etc.
Omosudis
Paralepis
Notolepis
Mautichthys
Lestidium
Lestidiops
Unasudis
Lestrolepis
Stemonosudis
Macroparalepis
Dolichosudis
Sudis
Ca. 300
3
2
1
18 +
2
3
3
2
2
18
■>
13
1
17
2
4
15
Ca. 30
1
2
1
7
1
5
3
1
4
20
4
3
13
7
1
2
G
G
7
7
H
H
7
H
H
H
H
H
H
H
H
H
G
G
G
G
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
+
(+)
( + )
( + )
( + )
( + )
( + )
( + )
+
+
+
+
+ + Moser and Ahlstrom (1970, 1974)
+ Okiyama (1974b)
+ Okiyama ( 1974b), Butler and Ahlstrom ( 1 976)
+
+ +
+
+ +
+ +
+ +
+ +
+ +
+
+
+ +
+ +
+ +
+ +
+ +
+ +
+ +
+ +
+ +
+ +
+ +
+
+ +
+ +
+ +
+ +
Tuning (1918)
Okiyama (1981)
Okiyama (1972), Parin and Belyamna (1972)
Okiyama (this study)
Sanzo (1938b). Okiyama ( 1 974b)
Bertelsenet al. (1976), Ozawa (1978)
Bertelsen et al. (1976), Ozawa (1978)
Bertelsen et al. (1976)
Johnson (1974b, 1982)
Marshall (1961), Rosen (1971). Johnson (1974a)
Okiyama (1979b)
Mito (1961a), Okiyama (1974b). Ozawa (1983)
Gibbs(1959), Okiyama (1974b), Ozawa (1983)
Okiyama (1974b)
Rofen (1966b)
Okiyama (this study)
Johnson (1982)
Ege (1958). Rofen (1966b), Belyanina (1981)
Ege (1930, 1957), Rofen (1966a)
Rofen (1966a)
Rofen (1966a)
Rofen (1966a)
Rofen (1966a)
Rofen (1966a)
Rofen (1966a)
Rofen (1966a)
Sanzo (1917). Rofen (1966a), Shores (1969),
Belyamna (1981)
■ For the details, see relevant section. ""G: gonochonsm; H: hermaphroditism,
early developmental stages is available at least for a single species.
Parentheses indicate information available for transparent ovanan eggs. '^ Double crosses mean that a series of
in Table 56. information on the reproduction and development
is abundant even for the deep-water species contrary to the
situation of about 20 years ago (Gosline et al., 1966). General
larval characteristics of this order were summarized by Ahl-
strom and Moser ( 1 976). Selected meristic characters including
many original data are given in Table 57.
Aulopidae (Fig. I09A-B).— This bottom-fish family is generally
considered the most primitive representative of the order. Its
systematics are inadequately known; at least seven nominal and
two undescribed species (Yamakawa, pers. comm.) occur in the
warm waters of the world except for the Indian Ocean.
Complete early life history series including egg stages are known
only for Aulopus japonicus (Okiyama. 1974b, 1980). Fragmen-
tary larval accounts are also available for some unidentifiable
species. Suggested dichotomy in the larval morphology in this
family (Okiyama, 1974b) is apparently wrong due to the erro-
neous identification of the early stages oi " Aulopus filamento-
sus" in Sanzo ( 1 938b) and TSning (1918), which are now ascribed
to Bathypterois of the Ipnopidae.
Eggs of .4. japonicus are spherical (1.18-1.14 mm in diame-
ter), pelagic, transparent, without an oil globule, and with ir-
regularly raised meshes on the chorion surface. Similar features
are not present in the matured ovarian eggs of A. filamentosus
( 1.36-1 .44 mm in diameter) with numerous oil globules (Sanzo,
1938b). The known larvae differ in gut structure, size of the
prominent pigment section and relative width of the slightly
narrow eyes. However, the followmg features are shared in com-
mon: single prominent peritoneal pigment section located at the
middle or slightly anterior region of the body; gently curved
208
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 57. Selected Meristic Characters of Myctophiform Genera.
Suborder and family
Genus
Dorsal
Anal
Pectoral
Pelvic
Branchiostegals
Vertebrae
Aulopoidei
Aulopidae
Aulopus
14-22
8-14
11-14
9
10-17
36-53
Myctophoidei
Mytophidae*
Diaphus, etc.
10-26
12-27
0-22
8
6-12
28-45
Neoscopelidae
Neoscopelus
11-13
10-13
15-19
8-9
8-9
30-31
Scopelengys
11-13
12-14
12-17
7-8
8
29-35
Solivomer
12-14
9-11
14-16
8
9-11
40-41
Chlorophthalmoidei
Chlorophthalmidae
Chlorophthatmus
9-13
7-11
15-19
8-9
8
40-50
Parasudis
10
8-9
17
9
8
38-39
Balhysauropsis
10-12
10-11
17-24
9
8-9
44-56
Ipnopidae
Ipnops
8-11
11-19
12-16
8
9-12
54-61
Balhytyphtops
11-13
13-17
12-15
8
14-17
62-66
Balhymicrops
8-10
9-15
9-10
7-8
8-10
65-80
Bathyplerois
12-16
7-13
13-22
8-9
10-14
49-65
Notosudidae
Ahliesurus
9-11
17-21
10-12
9
10
42-50
Scopelosaurus
9-13
15-21
10-15
9-10
10
53-67
Luciosudis
10-13
17-20
12-14
9-10
10
57-59
Scopelarchidae*
Scopelarchus. etc.
5-10
17-39
18-28
9
8
40-65
Synodontoidei
Bathysauridae
Bathysaurus
15-18
11-14
15-17
7-8
8-12
50-63
Harpadontidae
Harpadon
10-15
11-15
11-13
9
16-26
39-56
Sauhda
10-13
9-13
11-16
9
13-16
43-67
Synodontidae
Synodus
10-15
8-15
10-15
8
12-18
49-65
Trachinocephalus
11-13
14-16
11-13
8
14
54-58
Alepisauroidei
Alepisauridae
Alepisaurus
29-49
11-19
12-16
7-10
7
47-51
Anotopteridae
Anotoplerns
0
14-16
12-15
9-11
8
78-83
Evermannellidae*
Evermannelta. etc.
10-13
26-37
11-13
9
8
45-54
Omosudidae
Omosudis
9-12
14-16
11-13
8
8
39-41
Paralepididae
Paralepis
9-12
20-26
14-17
8
8
60-77
Notolepis
8-11
23-34
9-13
8-9
8
74-90
Maulichlhys
10-12
22-24
15-17
9
8
64-65
Lestidium
9-11
26-33
11-13
9
8
75-91
Lestidiops
8-13
25-35
10-13
6?
8
75-100
Uncisudis
10-11
25-31
11-13
9
8
75-79
Lestrolepis
9-11
31-44
10-12
8
8
82-98
Stemonosudis
7-12
29-50
10-13
8-9
8
84-121
Macroparalepis
11-14
21-32
10-12
9
8
80-110
Dolichosudis
10
36-37
11-12
9
8
101
Sudis
12-16
21-24
13-15
9
8
52-61
• For the details, see relevant section.
head profile; short fins; anus far fiDrward with wide preanai
interspace; anteriorly placed dorsal and pelvic fins. A size series
of A. japonicus reveals the gradual and direct development, with
scant pigmentation throughout the pelagic stages; melanophores
are restricted to the eyes and the caudal and postanal regions,
other than the peritoneal section which increases in size in older
larvae. Sequence of fin formation is C-D-A-P,-?,. Full ray com-
plements are visible at about 13.3 mm, but vertebral ossification
is delayed until about 20 mm, the smallest bottom specimen
available in my collection. Ontogeny of the upper jaw bones is
remarkable in possessing maxillary teeth (1-3) in larvae smaller
than 1 1 mm. Two supramaxillaries, peculiar to this family, are
ossifying in metamorphosed juveniles.
Myctophidae (see Moser, Ahlstrom. Paxton, this volume).
Neoscopelidae (Fig. 709C-D^. — Systematics of this deep-sea pe-
lagic and benthopelagic family are well understood (Butler and
Ahlstrom, 1976; Nafpaktitis, 1977), except for 5o/;vc)Wfr which
is restricted to the tropical Western Pacific. The remaining two
genera are known from the world oceans. Developing eggs are
unknown. Mature ovarian eggs of Neoscopelus macrolepidotus
(0.83-0.98 mm in diameter) contain a large single oil globule
of 0.39-0.61 mm (Maruyama, 1970). Advanced larval stages
have been described and illustrated for Neoscopelus sp. (Oki-
yama, 1974b) and two species of Scopelengys (Butler and Ahl-
strom, 1976). They are characterized by large fan-shaped pec-
toral fins, large head with blunt snout tip, small round eyes,
laterally compressed deep body, and an oval patch of melano-
phores in the peritoneum, distinct from the solid peritoneal
pigment sections of most other myctophiforms. All fins differ-
entiate rapidly with the possible sequence as P,-D-A-C-P,, full
counts being attained at a small size (less than 10 mm). Pig-
mentation is clearly difl^erent between the two genera. Scope-
lengys lacks the pigment patch lying along the dorsum of the
rectum in Neoscopelus. Scopelengys uniquely develops a hori-
OKIYAMA: MYCTOPHIFORMES
Table 58. Comparison of the Larval Characters Among Four Genera of the Ipnopidae.
209
Characters
Ipnops
Bathytyphlops
Bathymicrops
Bathypterois
Head profile
slung down; flat top
slightly slung down;
flat top
slung down; flat
top
slung down; flat top
Pectoral fin
bilobed; rays long
elongated; fan-shaped
elongated
elongated; fan-shaped
Gut size
short
short
long
long
Anus position; close to
pelvic fin
pelvic fin
pelvic fin;
slightly
anal fin
Anus-anal fin space
wide
wide
wide
narrow
Peritoneal pigment section
absent
single
absent
*numerous (12-20) or
absent
Body pigment (melanophores)
scant
scant
abundant
scant
Possible sequence of fin
P, C-A-D-Pj
P, C-A-D-P,
P,-C-A-D-P,
P,-C-A D-P,
formation
Transformation complete
ca. 42 mm SL
43-93 mm SL
70-90 mm SL
ca. 42-43 mm SL
' Details are mentioned in the text.
zontal pigment bar across the head. Small preopercular spines
are known only in Neoscopelus whereas a long snout is peculiar
to Scopelengys.
Chlorophthalmidae (Fig. ]09E-F).— Of Ihe three genera of this
benthic family, the cosmopolitan Chlorophthalmus is particu-
larly diverse and abundant. Extensive revision of this genus is
needed, since there are many undescribed species from the West-
em Pacific and the known species can be divided into two dis-
tinct groups, each warranting generic status (Doi and Okamura,
1983).
Eggs are not known. Despite the abundance of adults, few
larvae have been reported. Complete developmental series are
available for only C. agassizi (Tamng. 1918). Known larvae of
other species such as C. mento, C. prondens and Chlorophthal-
mus spp. (Pertseva-Ostroumova and Rass, 1973; Miller et al.,
1979; Okiyama, unpubl.) show close resemblance to C. agassizi
having the extremely short gut with large preanal interspace, a
similar pigment pattern composed of a single peritoneal pigment
section lying at the pectoral fin base and a melanophore at the
hypural complex, short fins and anteriorly placed dorsal and
pelvic fins (as in Aulopidae). There are possible specific differ-
ences in the size at appearance of the peritoneal pigment section
(ca. 7 mm in C. prondens vs 5-6.6 mm in C. mento) and in the
arrangement of the few small melanophores on the dorsal and
ventral margin of the tail near the notochord tip in early larvae.
Meristic characters are useful in discriminating the particular
species or species groups, although early developmental stages
are usually very difficult to identify to species.
Larval osteology was studied in detail for C. agassizi (Rosen,
1971) but the sequence of fin formation is not clear except that
the pectoral fin develops early. Principal changes during the
gradual metamorphosis include the rotation of the eyes dorsally
which takes place at sizes less than 40 mm (Ahlstrom, 1972a).
Unusual larvae with a pigmentation pattern similar to the
above described forms are found in ORl collections from the
Kuroshio area (Fig. 1 09 A). These are distinct in that the head
is markedly depressed, bowed with duckbilled appearance, and
a single peritoneal pigment section is large enough to cover the
dorsal half of the short gut. Their meristic characters (ca. 42
myomeres and ca. 1 7 pectoral rays) suggest a possible affinity
with Chlorophthalmus (sensu lato). These two larval types seem
to substantiate the suggested dichotomy of this genus. No in-
formation is available for larvae of the other two genera (Par-
asiidis and Bathysauropsis).
Ipnopidae (Fig. 1 1 OA-E). — Four benthic genera compose this
family which has been variously classified (e.g.. Nielsen. 1966;
Sulak. 1977). Despite their deep-sea mode of existence, larval
stages of all genera have been mostly obtained from the surface
waters. Developing eggs are not known. Mature ovarian eggs
are known for all genera with virtually identical features such
as a spherical shape, diameter of about 1.0-1.2 mm, and the
presence of a single large oil globule (Nielsen, 1966; Sulak, 1977
and pers. comm.; Merrett, 1980). Although intergeneric differ-
ences of the early larval stages are remarkable (Table 58). they
share several conspicuous characters including the more or less
hung-down head profile and the elongated precocious pectoral
fins. At metamorphosis these become less prominent in asso-
ciation with the drastic change in the mouth size from moderate
to huge and the appearance of heavy body pigmentation.
Two larvae (13.9, 10.6 mm) are known for Ipnops: the larger
specimen referred to /. agassizi was described in considerable
detail and illustrated (Okiyama, 1981). The smaller one may be
/. meadi in view of its higher anal ray count (ca. 13). A divided
pectoral fin with elongated upper rays is peculiar to this genus
(Table 58). Principal changes at metamorphosis include the de-
velopment of the unique eye plaque, a depressed head with
straight profile, and the disappearance of the peculiar feature of
the pectoral fins along with the loss of several rays. Metamor-
phosis may be rapid, but the smallest benthic juvenile of 40
mm still bears the immature eye plaque (Sulak, 1977).
Bathytyphlops includes only two species, B. sewelli and S.
marionae (Merrett. 1980). A larva of this genus was first de-
scribed under the name Macristiella perlucens of uncertain af-
finity (Berry and Robins, 1967). The known "Macristiella" (19
specimens, 7-43 mm) are all referable to B. marionae except
for the 37 mm larva from the Indian Ocean and the smallest
specimen (Parin and Belyanina, 1972). The Indian Ocean spec-
imen may be identified as B. sewelli on the basis of the higher
anal ray count (18), a unique character for this species.
Early stage larvae have little melanistic pigmentation, but
some bluish or violet coloration is present on the fins and var-
ious body parts in living specimens (Berry and Robins, 1967).
Preserved individuals sometimes retain this feature, usually on
the large pectoral or pelvic fins. Reduction of the relative size
of eyes, and the loss or replacement of the teeth as well as gill
rakers are among the major changes at metamorphosis, in ad-
dition to those common to the family. Otherwise, larval de-
velopment is rather direct and the relative position of the fins
and the anus changes little throughout ontogeny. The osteology
210
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
OKIYAMA: MYCTOPH I FORMES
211
Fig. 109. (A) Aulopus japonicus. 1 1.5 mm SL, from Okiyama (1974b); (B) Aulopus sp., 12.3 mm, from Okiyama (1974b); (C) Neoscopelus
sp., 7.9 mm, from southwestern Japan, Ocean Research Institute (ORI) collection; (D) Scopelengys dispar, 6.3 mm, from Okiyama (1974b); (E)
Chlorophlhalmus sp., 17.1 mm, from Indian Ocean, ORI collection; (F) Chlorophthalnms (?) sp., 7.5 mm, from Kuroshio waters off Japan, ORI
collection.
of both larvae and adults is well known (Okiyama. 1972; Parin
and Belyanina, 1972; Sulak, 1977).
Bathymicrops represents the deepest living myctophiform.
Two species. B. regis and B. brevianalis, are known from ex-
tremely limited material from 4225-5900 m (Nielsen, 1966;
Merrett and Marshall, 1981). Pelagic eggs are unknown. A total
of five larvae and juveniles (13.0-70.0 mm) are available; the
smallest two larvae (13.0, 14.7 mm) from Hawaiian waters are
unidentifiable; a 20 mm larva from the North Atlantic (=Sto-
miatella B in Roule and Angel, 1930: PI. 1, Fig. 7) is ascribed
to B. regis; the largest two juveniles (62.5, 70.0 mm) from the
tropical Pacific are tentatively identified as B. brevianalis.
Despite conspicuous variation among specimens, scattered
melanophore patches and an extremely slender body are diag-
nostic for this genus. The precocious pectoral fins are greatly
elongated even in the smallest larva, but the raised bases of the
dorsal and anal fins and the prominent finfolds are peculiar to
the advanced stages, which also have reduced eye size and a
slightly shorter gut. Size at metamorphosis is unusually large,
attaining 70-90 mm.
Bathypterois is the most speciose genus in this family. Three
subgenera (Benthosaurus, Bathypterois and Bathycygnus) and
18 species are currently included (Sulak, 1977). Known bathy-
metric ranges are 250-5,990 m. Published information of the
developmental stages is scant. Pelagic eggs are not known. A
single larva of 14. 1 mm (Okiyama, 1974b) was identified as B.
{Bathycygnus) longipes by Sulak (1977). As stated before, the
known early stages of "Aulopus filamentosus" are all referable
to those of Bathypterois. probably B. (Bathypterois) mediter-
raneus in view of their localities. Complete series of early stages
are confined to this species, but at least three additional larval
forms are now available. These known larvae share the distinct
forward shift of the ventral hypural elements in addition to the
features given in Table 58.
Known larvae are provisionally divided into two groups on
the basis of the peritoneal pigment sections, those with many
sections and those which lack peritoneal pigment. Except for
two larvae, B. (B.) longipes and B. (Benthosaurus) viridensis
(33.1 mm) from the Atlantic (Fahay, 1983), all specimens have
the former character state. The number of peritoneal pigment
sections can be a useful tool in discriminating the lai^ae, but
ranges of variation often overlap among species. A western Pa-
cific form with 12-18 pigment sections bears close resemblance
to B. (B.) mediterraneus larvae whereas decidedly lower myo-
mere counts of the former (45-48) readily separate these two.
B. viridensis larvae have, in addition to the complete absence
of the peritoneal pigment sections, several peculiar features such
as a slightly telescopic eye, a protruding gut, and a long anal fin
and short tail. Comparison with the smallest demersal specimen
(43 mm) of the same species (Sulak, 1977) indicates that prin-
cipal metamorphic changes include the absorption of the pro-
duced gut, lengthening of the posterior body and fin shrinkage.
This may represent the most pronounced metamorphosis in this
genus, since less remarkable transformation predominated in
the other species. Identification of the other larval types remains
to be determined.
Notosudidae (Fig. 1 J lA-B). — Bertelsen et al. ( 1 976) extensively
revised this oceanic midwater family, including information on
early developmental stages of all species (except Scopelosaurus
cradockei'). Supplemental information on the early stages is
available in Ozawa (1978). Pelagic eggs are unknown. Maturing
ovarian eggs of Ahliesaurus (ca. 0.3 mm in diameter) and Lu-
ciosudis (0.4-0.5 mm) suggest that pelagic eggs are uncommonly
small for this order.
General characteristics of these larvae are extremely similar
throughout the family: long, slender subcylindrical body, be-
coming increasingly compressed toward the tail; markedly de-
pressed head with wedge-like snout; posteriorly protruding lobes
in corpus cerebelli; narrow eye with longer horizontal axis; a
more or less distinct conical mass of choroid tissue on the pos-
terior part of slightly stalked eye; anus at about midbody (except
.4hliesaurus) widely separated from anal fin origin; slight in-
crease of gut length with growth during the early larval stages;
absence of the peritoneal pigment. Maxillary teeth peculiar to
larvae help diagnose this family but are not unique (see, Au-
lopidae). Possible sequence of fin formation is CA-D-P.-Pj,
212
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
last elements being rarely visible in larvae less than 20 mm.
Apart from the length at metamorphosis varying between 25
and 45 mm among species, pigmentation pattern is usually the
only useful character for specific identification. Once established
these pigment patterns, mostly restricted to the tail, are retained
throughout the larval stages, although a few species are known
to be unpigmented throughout all or part of the larval period.
Scopelarchidae (see R. K. Johnson, this volume).
Bathysauhdae (Fig. 1 1 IC).— This deep-water benthic family
consists of two species of synchronous hermaphrodites, Bathy-
saurus mollis and B. ferox (=B. agassizi) (Sulak, pers. comm.;
Wenner, 1978).
Pelagic eggs are unknown. Maximum size of mature ovarian
eggs in B. ferox is 1.2 mm in diameter (Wenner, 1978). So-
called "Macristium" forms are now proved to be larval Bath-
ysaunis (Rosen, 1971; Johnson, 1974a); at least several of the
five known "Macristium" larvae (20-83 mm) are positively
identified with B. mollis. Morphology and osteology of these
specimens have been closely studied, revealing many charac-
teristic features such as unusually elongated fins, anterior place-
ment of dorsal and pelvic fins, raised bases of dorsal and anal
fins, long gut (coiled or uncoiled) terminating just in front of
anal fin origin, six peritoneal saddle-shaped pigment sections
all evenly spaced, and development of a pattern of lateral bars
in some specimens. Besides this last feature, meristic differences
serve to distinguish two species despite considerable variation.
Metamorphosis may take place gradually at exceptionally large
sizes (more than 83 mm). Accompanying changes include short-
ening of fins, expansion of the gape with necessary associated
changes in head bones and associated anatomy, backward shift
of the dorsal fin origin, and darkening of the body surface, oral
cavity and peritoneum.
Harpadontidae (Fig. 1 1 ID-E). — Two genera are recently in-
cluded here (Sulak, 1977; Johnson, 1982). Harpadon comprises
at least four species living in nearshore waters, estuarine and
relatively deep continental shelf waters of the Indo-Pacific. Crit-
ical systematic revision of this genus is now in progress (Schmitz,
pers. comm.). A pelagic egg referred to H. nehereus in Delsman
( 1929c) appears invalid (Delsman and Hardenberg, 1934). Early
developmental stages are poorly studied; only two specimens
of//, nehereus (25.2, and ca. 40 mm) have been illustrated and/
or briefly described (Delsman and Hardenberg, 1934; Okiyama,
1979b). A juvenile of 55 mm is the smallest specimen of the
deep water congener, //. microchir, available in ORI collections.
Early stages are readily discriminated from most other myc-
tophiform larvae by the exceptionally high numbers of bran-
chiostegal rays (16-27) and the following characters: elongate
compressed body with large head and mouth, short snout (due
to the forward shift of eyes), scant pigmentation except seven
pairs of peritoneal pigment sections, the last two closer together
than the others, and extension of the lateral line scales onto the
caudal fin. Of these rather advanced developmental features,
pigmentation pattern may be common to the earlier stages.
Apparently, long pectoral and pelvic fins are peculiar to //.
nehereus. Also, //. microchir is more lightly pigmented than //.
nehereus at similar lengths.
Metamorphosis seems gradual. If the occurrence of melano-
phores over the stomach is of significance in defining this pro-
cess, transformation is completed by 35 mm in //. nehereus.
There are about 1 5 species of Saurida with highest diversity
in the Western Pacific. Planktonic eggs are known for S. elon-
gata, S. wanieso. and S. tumbil besides several unidentifiable
species (Mito, 1961a; Zvjagina, 1965a; Venkataramanujan and
Ramanoorthi, 1981). These are spherical, 1.0-1.3 mm in di-
ameter, transparent, without oil globules and with a narrow per-
ivitelline space. Hexagonal sculpturing on the chorion (0.03-
0.05 mm in mesh size) is either present (S. wanieso and S.
tumbil) or absent (S. elongata). Early developmental stages are
known for 9 species. Of these, complete developmental series
are available for at least 4 Pacific species, S. tumbil, S. elongata.
S. wanieso and S. gracilis (Dileep, 1977; Ozawa, 1983) and the
Atlantic species, 5. brasiliensis (K\x(i.omtX]f.ma.. 1980). These lar-
vae are extremely similar to those of Harpadon. except for the
lower numbers of branchiostegals and invariably short fins.
Complete absence of the preanal finfold in the early stages is
peculiar to this genus (Ozawa, 1983). Except for S. brasiliensis.
however, these are divided into two types on the basis of pig-
mentation pattern. One of these consisting of S. gracilis and
probably some Atlantic congeners is characterized by evenly
spaced peritoneal pigment sections of similar size and simul-
taneous differentiation. In addition, prominent pigment along
the anal fin base and on the caudal fins may be diagnostic for
this type. 5. gracilis larvae uniquely develop a small choroid
mass on the ventral side of narrow eyes (Ozawa, 1983) while
nothing is mentioned in this regard for Hawaiian larvae (Miller
et al., 1979). Remaining larvae belong to the second type in
which the terminal pigment section is smaller and later-ap-
pearing than the anterior sections. Other pigment is also scarse
or absent in this latter type, where specific differences are known
in the size of pigment sections and vertebral numbers. Meta-
morphosis occurs fairly gradually with considerable variation
in size among species, but is usually complete before 40 mm
(Gibbs, 1959).
Fig. 110. (A) Ipnops agassizi. 13.9 mm SL, from Okiyama (1981); (B) Balhytyphlops manonae. 13.1 mm, from Okiyama (1972); (C)
Bathymicrops brevianalis. 70.0 mm, from tropical central Pacific, ORI collection; (D) Bathypterois sp. (pigmented type), from northeast of
Australia, Southwest Fisheries Center (SWFC) collection; (E): Bathypterois viridensts (unpigmented type), from Fahay (1983).
Fig. 111. (A) Scopelosaurus smilhii. 1 3.4 mm SL, from southwestern Pacific, ORI collection; (B) the same, dorsal view of head; (C) Bathysaurus
ferox. 33.0 mm, from Marshall (1961); (D) Harpadon nehereus. 25.2 mm, from East China Sea, ORI collection; (E) Saunda undosquamis. 15.6
mm, from Okiyama (1974b); (F) Synodus lucioceps. 10.5 mm, from California current region. SWFC collection; (G) Trachinocephatus myops.
21.3 mm, from Zvjagina (1965a).
Fig. 112. (A) Atepisaurus brevirostris. 12.1 mm, from Rofen(1966b);(B)/l./era>:. 10.0 mm, from central Pacific near Hawaii, SWFC collection;
(C) Anotopterus pharao, 14.2 mm, from California current region, SWFC collection; (D) Omosudis lowei (central western Atlantic specimen),
11.8 mm, from Rofen (1966b); (E-F) O. lowei. 22.5 mm, from tropical western Pacific, ORI collection, showing dorsal view of head.
OKIYAMA: MYCTOPHIFORMES
213
214
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
OKIYAMA: MYCTOPHIFORMES
215
216
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Synodontidae (Fig. 1 1 IF-G). — Synodus includes about 30 species
and has a circumglobal distribution with distinctly high diversity
in the Indo-Pacific. Another monotypic genus of this family
(Trachinocephalus) shows world-wide distribution. A recent re-
vision of the Indo-Pacific Synodus (Cressey, 1981), including
many new species, critically changed its systematic status. Thus,
most of the known eggs and larvae are subject to nomenclatural
revision. Early stages of this family can be separated from those
of the previous family by the presence of the preanal finfold
(Ozawa, 1983).
Trachinocephalus myops larvae are distinct in possessing six
pairs of large peritoneal pigment sections of uniform size, a
rounded head with short snout, and additional unique pigmen-
tation (Rudometkina, 1980; Ozawa, 1983). This species and
most species of Synodus have an extremely elongated body. An
exception is the eastern Pacific species, S. lucioceps. which has
a slightly deeper body. A complete developmental series is known
only for this species in Synodus; eggs are spherical, 1.33-1.44
mm in diameter, without an oil globule, with moderately broad
perivitelline space and hexagonally sculptured chorion: larvae
are characterized by 7 evenly spaced pairs of pigment sections
formed gradually, a ventral melanophore lying at the midpoint
of tail, and one near the notochord tip.
As in the Harpadontidae, meristic characters and pigmenta-
tion patterns are of particular aid in identifying the early stages
of this family. If established pigmentation patterns are retained
in the metamorphosed juveniles or adults, numbers of the per-
itoneal pigment sections of all Indo-Pacific species of Synodus
(Cressey, 1981) vary between 0 and 17 with a maximum range
of infraspecific variation of 0-3 in 5. binotalus and 14-17 in S.
usitatus; some species appear to lack this pigment (i.e., S. kaian-
us and S. binotalus), however this needs to be documented by
complete developmental series. Another point of interest is the
asymmetry and size disparity of the pigment pairs known in
"S. variegatus" of Okiyama (1974b).
Size at metamorphosis and sequence of fin formation of this
family appear to be identical to those in Harpadontidae. Ozawa
(1983) revealed the following pattern of fin formation: C-A-D-
P,P2.
Alepisauridae (Fig. //2,4-BA— This widely distributed bathy-
pelagic family includes only two species, .Hepisaurus fero.x and
A. brevirostris. with slightly different ranges: the latter is appar-
ently absent from the North Pacific (Francis, 1981). Eggs are
unknown. A series of early developmental stages of Alepisaurus
sp. (6.9-17.2 mm) has been described and illustrated (Rofen,
1966b). In addition, three larvae (9.6-16.5 mm) from the col-
lection of the Southwest Fisheries Center, La Jolla have different
features. They share with previous specimens a large head and
mouth, prominent canine teeth on the dentary, small fins in-
cluding pigmented pectorals of moderate size, gently curved
head profile and short gut with heavy pigmentation. The peri-
toneal pigment section is indistinct. This new material is unique
in having 4 small preopercular spines, pigment patches at the
anal fin origin, and distinct bony ridges dorsally on the head.
Judging from the locality of these specimens, near Hawaii in
the North Pacific, Alepisaurus sp. larvae of Rofen (1966b) can
be identified with A. brevirostris. and these with A.ferox.
Metamorphosis may be gradual with possible sequence of fin
formation P, C-D-A-Pj.
Anotopteridae (Fig. 112C).— One world-wide species, Anotop-
terus pharao, constitutes this open ocean family, uniquely lack-
ing the dorsal fin. Eggs are not known. A larva (ca. 1 5 mm) has
been briefly described without illustration (Nybelin, 1948): this
specimen is unavailable now (Thulin, pers. comm.). Another
larva of similar size (14.2 mm) is available from the collection
of the Southwest Fisheries Center, La Jolla. It is characterized
by a slender thin body, absence of peritoneal pigment sections,
large head with pointed snout, a fleshy prolongation at the tips
of both jaws, two large canine teeth on each palatine, and a
fairly long gut extending beyond midbody. Pigmentation is scat-
tered on various parts of body including the snout, jaw tips,
dorsal midline of body, near the tail tip, and peritoneum (par-
ticularly along the dorsum of gut). Except for the pectoral fin,
fin aniages are lacking. A juvenile of about 50 mm illustrated
in Rofen ( 1 966c) is similar to the described larva, except all fins
are differentiated including the adipose fin: body pigmentation
is remarkable in this juvenile. Perhaps, this species has the most
direct pattern of early development in this order.
Evermannellidae (see R. K. Johnson, this volume).
Omosudidae (Fig. ] 12D-F).—A single mesopelagic species,
Omosudis lowei. constitutes this cosmopolitan family. Pelagic
eggs are not known. Excellent developmental series have been
described and illustrated, chiefly based on Atlantic material
ranging from 5.7 to 75.2 mm (Ege, 1958: Rofen, 1966b). Re-
cently, a larva (11.5 mm) with different features was briefly
described and illustrated (Belyanina, 1982b). Its locality in the
tropical western Pacific is peculiar and additional specimens are
available in ORI collections (pers. obs.).
These have in common a very large head and mouth, stubby
body, long pointed snout, straight head profile, small fins, par-
ticularly the pectoral, large canine teeth on denlary and palatine,
and several closely spaced peritoneal pigment sections. How-
ever, trenchant morphological differences between the Atlantic
and Pacific specimens are known: head smooth vs armed (along
edge of preopercle and dorsum of head): pigmentation light vs
dense at a similar size: pigmented band above posterior part of
anal fin absent vs present. For this first character, there is a
possibility that the minute preopercular spines have been over-
looked in the Atlantic larvae.
Sequence of fin formation known in the Atlantic specimens
is C-DA-Pj-P,. Metamorphosis is gradual with possible dif-
ferences in the size of completion between the two types as
suggested above. The presence of two larval types is in sharp
contrast with the current concept of a monotypic family. In this
connection, Ege's comments ( 1 958) on the significant differences
in dorsal ray numbers between the populations from the South
China Sea and north Atlantic are of particular interest.
Paralepididae (Fig. 1 13. 4-G}.— This oceanic pelagic family in-
cludes about 1 1 genera and 50 species and constitutes the second
largest group in the order after Myctophidae. Some genera are
still in need of critical revision, while the two established
subfamilies seem valid. Paralepidiinae includes two tribes, the
Paralepidiini (3 genera) and Lestidiini (7 genera), and Sudinae
has I genus (Sudis). Ege (1930) and Rofen (1966a) mcluded
early larval stages in their extensive studies of this family. Eggs
are not known but developmental stages are known for 9 out
of 1 1 genera. Larval development of Sudis has been closely
studied for 5. hyalina and S. a/ro.v (Sanzo, 1917: Shores, 1969:
Belyanina, 1981). These unusual larvae are readily discrimi-
nated from those of the other subfamily by the relatively short
body with large head, long pectoral fins, long gut and early
OKIYAMA: MYCTOPHIFORMES
217
Fig. 113. (A) Paralepis elongata. 16.7 mm SL, from Rofen (1966a); (B) Notolepis coatsi. 60.5 mm, from Efremenko (1983); (C) Leslidiops
ringens. 9.4 mm, from California current region. SWFC collection; (D) the same, 28.5 mm; (E) Stemonosudis macrura. 1 1.2 mm, from Ege(1957);
(F) Sudis hyalina. 16.1 mm, from Shores (1969); (G) 5. alrox. 21.5 mm, from Berry and Perkins (1966).
established complement of peritoneal pigment sections, spine-
tipped flanges on ventral region of preoperculum. over eye, and
snout. 5". alrox has a spine-tipped flange along lower jaw. and
the large spine at the preopercular angle is serrated only in 5.
airo.x. The precocious pectoral fin is relatively short until about
1 5 mm in S. atrox whereas it is very long even in 8.0 mm larvae
of 5. hyalina. The number of peritoneal pigment sections is 6
(5 in early larvae) in S. atrox vs 7-8 in S. hyalina. Trunk
pigment is evenly distributed in S. atrox \s patchy in S. hyalina.
Except for this genus, the developmental features of this fam-
ily are remarkably cohesive. Known lar\ae have a very long
compressed body, a short trunk in early larvae, large head in
advanced larvae, elongated pointed snout with straight head
profile, various numbers of peritoneal pigment sections sequen-
218
ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM
tially formed with gradual lengthening of gut, well developed
preanal finfolds and apparently precocious anal fin rays. Ad-
ditionally, during ontogeny eye shape changes from ovoid to
round, and body pigmentation changes from light to dense.
These larvae are too similar in general appearances to determine
trenchant characters that define genera or tribes. Peritoneal pig-
ment sections, are of prime importance in identifying early stages,
but show extreme variability with respect to their number and
sequential development. Of particular interest in this connection
is Notolepis. N. rtssoi develops 1 2 pigment sections, the largest
number in the family except Stemonosudis (3 1 ), whereas the
Antarctic congener, TV. coalsi, has only a single section which
increases in size with growth (Efremenko, 1978, 1983a). Among
the various genera the primary section develops at 5-10 mm
and full complements are formed variously by the species be-
tween 15-45 mm. Usually, metamorphosis takes place around
this size accompanied by the development of a black perito-
neum.
In addition to the exceptionally higher number of pigment
sections, Stemonosudis is peculiar in having a filamentous pro-
jection on the lower jaw tip (in larvae of 5. macrura and in
juveniles and adults of S. intermedia and 5. elongatd). Likewise,
Uncisudis (=Pontosudis) uniquely develops an elongated pelvic
fin.
Patterns of melanophores are extremely diverse but of use in
identifying species or species groups; pigment patches on the
caudal peduncle, dorsum of body, and caudal and pectoral fins
are particularly important. Rofen (1966a) suggested that the
single larval character discriminating the two tribes in Parale-
pidiinae, i.e., Paralepidiini and Lestidiini, is whether the rear-
ward shift of the anus occurs early or late in ontogeny.
Incertae ce^w. — Peculiar eggs described by Delsman (1938) and
Mito (1961a) are currently considered to be those of mycto-
phiform fishes other than Myctophidae (Moser and Ahlstrom,
1970). These eggs are spherical, 1.12-1.37 mm in diameter, with
a single oil globule and bear numerous short appendages on the
chorion. Two types are known only from Asian waters.
Ocean Research Institute, University of Tokyo,
MiNAMiDAi, Nakano-ku, Tokyo 164, Japan.
1-15-1,
Myctophidae: Development
H. G. Moser, E. H. Ahlstrom and J. R. Paxton
LANTERNFISHES of the family Myctophidae are found in
all oceans of the world. Some 230-250 species are arranged
in 36 generic/subgeneric taxa (Table 59). All nominal species
are listed in Paxton (1979). Characteristic of the family is the
presence of light organs or photophores on the head and body
(Fig. 1 1 4). The different patterns of photophores have been used,
along with meristics (Table 60), in species diagnoses and as a
basis for classification within the family since the late 1800's.
Most authors have placed the Myctophidae and closely related
Neoscopelidae with the families Aulopidae, Chlorophthalmidae
and related families in an order or suborder variously named
the Iniomi, Myctophoidea or Myctophiformes (Gosline et al.,
1966; Greenwood et al., 1966; Nelson, 1976; Johnson, 1982),
although Rosen ( 1973) separated the Myctophidae and Neosco-
pelidae as a restricted order Myctophiformes. Moser and Ahl-
strom (1970, 1972, 1974), Ahlstrom et al. (1976) and Paxton
(1972) are the most recent papers considering relationships with-
in the family; characteristics of larvae and bones and photo-
phores of adults were primarily utilized in the respective studies.
Paxton's (1972) classification, including genera recognized sub-
sequently, is as follows:
Subfamily Myctophinae
Tribe Electronini
Genera: Protomyctophum.
Metelectrona-
Krefftichlhys', Elect rona.
Tribe Myctophini
Genera: Benthosema, Diogenichlhys, Hygophum, Myc-
tophum. Symbolophorus
Tribe Gonichthyini
Genera: Loweina, Tarletonbeania, Gonichthys, Centra-
branch us
Subfamily Lampanyctinae
Tribe Notolychnini
Genus Notolychnus
Tribe Lampanyctini
Genera: Taaningichthys, Lampadena, Bolinichthys. Lep-
idophanes, Ceratoscopelus. Stenobrachius, Lampan-
yctus, Triphoturus, Parvilux^
Tribe Diaphini
Genera: Lobianchia, Diaphus, Idiolychnus*
Tribe Gymnoscopelini
Genera: Lampanyctodes, Gymnoscopelus, Notoscopelus,
Lampichthys, Scopelopsis, Hintonia
There has not been a family revision at the species level since
Fraser-Brunner's (1949) study. A large number of more recent
generic revisions and regional studies are currently the primary
sources for species identifications; most of these have been uti-
lized in compiling the generic distribution limits (Table 59). The
most recent zoogeographic studies are those of Backus et al.
Hulley (1981).
■ Wisner(1963).
' Hubbs and Wisner (1964).
" Nafpaktitus and Paxton (1978).
MOSER ET AL.: MYCTOPHIDAE
219
Table 59. Geographic Distribution of the Genera and Subgenera of Myctophidae. References marked * are useful for the identification
of species. The division of the Atlantic and Indian Oceans is arbitrarily taken at 20°E, the Indian-Pacific Ocean boundary at 130°E.
No, of
species
Lai. extremes
Krefftichthys
I Atlantic
Indian
Pacific
34°S-60°S
43°S-66°S
34°S-72°S
Protomyclophum
(Protomyctophum)
7 Atlantic
Indian
Pacific
34°S-60°S
44°S-65°S
40'^70°S
Protomyctophum
(Hierops)
7 Atlantic
Indian
Pacific
70°N-56°S
35°S-52°S
57°N-67°S
Electrona
5 Atlantic
Indian
Pacific
55°N-70°S
2°N-68°S
42°N-70°S
Metelectrona
2 Atlantic
Indian
Pacific
35''S-5I°S
35°S-47°S
33°S-55°S
Benthosema
5 Atlantic
80°N-38°S
Indian
Pacific
2rN-35°S
7I''N-42°S
Diogenichthys
3 Atlantic
Indian
Pacific
50°N-48°S
18°N-45°S
37°N-41°S
Hygophum
9-1 1 Atlantic
Indian
Pacific
49°N-48°S
20''N-42°S
39"'N-46°S
Symbolophorus
7-9 Atlantic
Indian
Pacific
59''N-51°S
2I°N-41°S
50°N-59°S
Myctophum
13-14 Atlantic
Indian
65''N-40°S
20°N-34°S
Pacific
42''N-42°S
Loweina
3-4 Atlantic
Indian
Pacific
44°N-38°S
IO°S-40'^
32''N-40°S
Tarletonheania
1-2 Atlantic
Indian
Pacific
50°N-30°N
Gonichthys
3-4 Atlantic
Indian
Pacific
47°N-40°S
25°S-39'>S
3I''N-42°S
Cenlrobranchus
3-4 Atlantic
Indian
Pacific
46°N-35°S
15°N-33°S
37°N-37°S
Nololychnus
I Atlantic
Indian
Pacific
56''N-38°S
1 rN-40°S
34'>N-44°S
Lobianchia
2 Atlantic
Indian
Pacific
6rN-5l°S
2'>N-40°S
32°N-47°S
Diaphus
65-75 Atlantic
Indian
Pacific
62''N-52'^
23''N-48°S
55°N-58°S
Idiolychnus
1 Atlantic
Indian
Pacific
13°S-24°S
2rN
*Hulley (1981:12)
*Hulley (1972:217); Andriashev (1962:224)
Andnashev (1962:225): McGinnis (1982:1 1)
*Hulley (1981:29, 19)
Hulley (1972:218); *McGinnis (1982:17)
♦Andriashev (1962); *McGinnis (1982:16, 17)
Nafpaktitis et al. (1977:31); *Hulley (1981:36)
•Nafpaklitis and Nafpaktitis ( 1 969:7); 'McGinnis (1982:18)
♦Wisner ( 1 976:20); 'McGinnis ( 1 982: 1 8)
*Hulley (1981:40, 46); *McGinnis (1982:21)
Nafpaktitis and Nafpaktitis (1969:10); *McGinnis (1982:21)
•Andriashev ( 1 962); Ebeling ( 1 962: 1 40); *McGinnis (1982:21)
•Hulley (1981:53)
•McGinnis (1982:25)
•Bussing (1965:200); •McGinnis (1982:25)
•Nafpaktitis et al. (1977:52); Hulley (1972:220); (the specimen from
55°S is possibly mislabeled, McGinnis, (1982:26, 29))
Kotthaus (1972:18); *Nafpaktitis and Nafpaktitis ( 1969:1 1)
•Wisner (1976); Nafpaktitis et al. (1977:52); Robertson et al. (1978:302)
Nafpaktitis et al. (1977:58); Hulley (1981:58)
•Nafpaktitis and Nafpaktitis (1969:15)
•Wisner (1976:49); Rass (1960:149)
•Bekker(1965); •Nafpaktitis et al. (1977:38); •Hulley (1981:61)
•Bekker (1965:80); Hulley (1972:222)
•Wisner (1976); •Bekker (1965:94); McGinnis (1982:30)
•Hulley (1981:101)
Kotthaus (1972:27); *Nafpaktitis and Nafpaktitis (1969:29)
•Wisner (1976); Frost and McCrone (1979:755); •McGinnis (1982:33)
•Nafpaktitis et al. (1977:62); 'Hulley (1981:87)
Nafpaktitis and Nafpaktitis (1969); •Bekker and Borodulina
(1978:1 20); McGinnis ( 1 982:34)
•Kawaguchi and Aioi (1972); •Wisner (1976); Kawaguchi et al.
(1972:27); Paxton and Nafpaktitis (ms)
•Nafpaktitis et al. (1977:85)
•Bekker (1964:23); •Nafpaktitis and Nafpaktitis (1969:31)
•Wisner (1976); •Bekker (1964:23); McGinnis (1982:37)
•Bekker (1963:160); 'Wisner (1976:82)
Nafpaktitis et al. (1977:88); Hulley (1981:107)
•Bekker (1964:38)
•Bekker (1964); •Wisner (1976:86); McGinnis (1982:36)
•Nafpaktitis et al. (1977:91)
•Bekker (1 964:5 1, 58)
•Bekker (1964:58)
•Nafpaktitis et al. (1977:94); 'Hulley (1972:222)
Kotthaus (1972:30); McGinnis (1982:37)
Ebeling (1962:141); McGinnis (1982:37)
•Nafpaktitis el al. (1977); Bekker (1967:98); McGinnis (1982:51)
•Nafpaktitis (1978:7); McGinnis (1982:51)
'Wisner (1976:96); McGinnis (1982:51)
'Nafpaktitis et al. (1977:158); McGinnis (1982:52)
'Nafpaktitis (1978:62, 78)
'Nafpaktitis (1978:62); McGinnis (1982:52)
'Nafpaktitis and Paxton (1978:495)
'Nafpaktitis and Paxton (1978:495-496)
220
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
Table 59. CoNTrNUED.
No of
species Ocean
Lai. exlremes
References
Lampanyctodes
1 Atlantic
Indian
Pacific
I9°S-34°S
35''S
34°S-5I°S
Gymnoscopelus
(Gymnoscopelus)
4 Atlantic
Indian
Pacific
34°S-66°S
60°S-65°S
40°S-72°S
Gymnoscopelus
(Nasolychnus)
4-5 Atlantic
Indian
Pacific
34°S-57°S
24°S-65°S
40°S-70°S
Scopelopsis
1 Atlantic
Indian
Pacific
11°S-48°S
9»S-40°S
15°S-35°S
Lampichthys
1 Atlantic
Indian
Pacific
30°S-48°S
35'^-40°S
7°S-49°S
Notoscopelus
(Noloscopelus)
5 Atlantic
Indian
65°N-60°S
8''S-36'^
Pacific
50°N-37°S
Notoscopelus
(Parieophus)
1 Atlantic
Indian
Pacific
50°N-2rN
Hintonia
1 Atlantic
Indian
Pacific
39°S-48''S
47<«-51°S
40°S-50°S
Lampadena
(Lampadena)
8-9 Atlantic
Indian
Pacific
65°N-48°S
6''N-49°S
4rN-49°S
Lampadena
(Dorsadena)
I Atlantic
Indian
Pacific
45°N
Taaningichthys
3 Atlantic
Indian
Pacific
43°N-44°S
8°N-30°S
4rN-68°S
Ceratoscopelus
3 Atlantic
Indian
52°N-45°S
20''N-43''S
Pacific
43°N-42°S
Lepidophanes
2 Atlantic
Indian
Pacific
43''N-48°S
Bolinichthys
7 Atlantic
Indian
Pacific
53°N-41°S
21°N-44°S
31°N-43°S
Triphoturus
3-4 Atlantic
Indian
Pacific
8°N-I4°S
38°N-35°S
Stenohrachius
2 Atlantic
Indian
Pacific
57<'N-30°N
Parvilux
2 Atlantic
Indian
Pacific
40''N-14°S
Lampanyctus
40 Atlantic
Indian
65°N-60°S
16°N-60°S
Pacific
59''N-72°S
♦Ahlstrom et al. (1976:146); Grindley and Pennth (1965:283)
Paxton and Nafpaktitis (in prep.)
*Wisner (1976: 1 58-1 59); McGinnis (1982:55)
*Hulley (1981:254); •McGinnis (1982:59)
*Andriashev (1962:267); •McGinnis (1982:59)
♦McGinnis (1982:61, 58)
♦Hulley (1981:261); (03°S, Fraser-Brunner (1931:224) presumably a
waiO
Smith (1 933a: 1 26); *McGinnis (1982:64)
♦Andriashev (1962); McGinnis (1982:64)
*Hulley (1981:241)
Legand (1967:49); McGinnis (1982:57)
*Wisner (1976:222); Paxton and Nafpaktitis (in prep.)
Hulley (1981:242)
McGinnis (1982:57)
*Wisner (1976:215); McGinnis (1982:57)
♦Nafpaktitis et al. (1977:254) Andriashev (1962:278)
Nafpaktitis and Nafpaktitis (1969:35); Grindley and Penrith
(1965:283)
*Fujkii and Uyeno (1976); Frost and McCrone (1979:755); Collins
and Baron (1981:11)
•Nafpaktitis et al. (1977:257)
•Hulley (1981:239)
McGinnis (1982:55)
•Wisner (1976:220); McGinnis (1982:55)
•Kreflt (1970:285); Hulley (1981:180)
•Nafpaktitis and Paxton (1968:20, 21)
•Nafpaktitis and Paxton (1968:20, 21)
•Coleman and Nafpaktitis (1972:2)
•Hulley (1981:167); 'Davy (1972)
•Nafpaktitis and Nafpaktitis (1969:40)
•Davy (1972:72); •Nafpaktitis et al. ( 1 977: 1 9 1 )
•Nafpaktitis et al. (1977:243); Hulley (1981:237)
•Bekker and Borodulina (1968:792); •Nafpaktitis and Nafpaktitis
(1969:65)
•Wisner (1976:207); Robertson et al. (1978:302)
•Nafpaktitis et al. (1977:225); •Hulley (1981:223)
•Nafpaktitis et al. (1977:240); •Hulley (1981:229)
Kotthaus (1972:18): 'Nafpaktitis and Nafpaktitis (1969:60)
•Johnson (1975:58); Nafpaktitis et al. (1977:234)
Hulley (1981:205)
•Nafpaktitis and Nafpaktitis (1969:51)
•Wisner (1976:165)
•Wisner (1976:160)
•Wisner (1976:163, 164)
•Nafpaktitis et al. (1977:196); •Hulley (1981:183); Zahuranec (1980)
•Nafpaktitis and Nafpaktitis (1969); Kotthaus (1972:35); •McGinnis (1982:42);
Zahuranec (1980)
•Wisner (1976:191); McGinnis (1982:42); Zahuranec (1980)
MOSER ET AL.: MYCTOPHIDAE
221
Table 60.
Meristics of the Genera
AND Subgenera of
Myctophidae.
Fin rays
Branchio-
Dorsal
Anal
Pectoral
Pelvic
Procurrenl caudal
Vertebrae
stegals
Gill rakers
Krefflichthys
11-14
17-19
14-16
8-9
8-9 + 7-9
36-39
6-8 + 19-23
Protomyctomphum
10-14
21*-27
14-17
8-9
7-9 + 6-9
35-41
8-10
4-7 + 14-21
P. Hierops
11-13
20-27
15-18
8
7-1 1 + 6-9
36-42
9-10
3-5 + 13-18
Electrona
12-16
18-22
11-17
8
6-10 + 6-9
33-41
7-8
3-10 + 12-25
Metelectrona
13-15
19-22
14-16
8
10 + 9
35-38
8
4-7 + 16-20
Benthoscma
11-15
16-22
10-17
8-9
7-9 + 7-9
31-37
9
3-10 + 10-21
Dwgenichlhys
10-13
14-18
10-14
7-8
7-9 + 7-9
29-34
7
2-4 + 10-12
Hygophum
10-15
18-25
12-17
8-9
6-9 + 6-9
34-40
9
3-6 + 12-16
Myclophum
11-15
16-27
12-22
7-8
7-9 + 7-9
35-46
8-9
4-8 + 10-21
Symhotophorus
12-16
18-24
12-20
8
8-10 + 7-9
36-42
9
4-7 + 12-19
Loweina
10-13
13-17
9-12
7-9
6-7 + 6-7
37-39
9
2-3 + 5-10
Tarlelonheania
11-15
16-20
11-16
8
5-8 + 5-8
40-42
8
4-6 + 10-12
Gonichlhys
10-13
17-24
11-18
6-8
5-6 + 5-6
38-41
9
3-6 + 7-12
Cenlrobranchus
9-12
16-20
11-17
8
5-7 + 5-7
35-40
7-8
0
Nololychnus
10-12
12-15
11-15
6-7
7-9 + 7-9
27-31
9-10
2 + 8-9
Lnlnanchta
15-18
13-15
11-15
8
5-7 + 5-6
33-35
9
4-6 + 11-16
Diaphus
10-19
11-19
9-14
8
5-8 + 5-8
31-37
8-9
4-11 + 9-21
Idwlychnus
14-15
14-16
13-15
8
34
6-7 + 14-15
Lampanyctodes
13-14
14-17
12-14
8
8-10 + 9-10
36-39
9-11
10-11 + 20-23
Gymnoscopelus
14-21
16-22
12-16
8-9
10-12 + 11-15
41-45
10
6-12 + 14-26
G. Nasolychnus
16-20
16-20
12-15
8
8-13 + 10-15
41-45
10-11
7-12 + 17-25
Scopelopsis
20-24
23-27
10-12
7-8
9-11 + 11-12
38-39
9-10
7-9 + 16-18
Lainpkhlhys
16-18
21-23
11-15
8
10 + 12
40-41
9
4-6 + 13-16
Noloscopelus
21-27
18-21
11-14
8-9
10-14 + 10-15
35-40
10
4-10 + 9-22
N. Parieophus
23-26
18-20
12-14
37-38
8-10 + 18-20
Hmtonia
14-16
12-14
13-15
8
10-11 + 13
37-39
9
6-7 + 11-14
Lampadcna
13-16
12-15
13-18
8
8 + 8-9
35-40
9
3-8 + 9-18
L. Dorsadena
14-15
12-14
15-16
8-9
4-5 + 12
Taaningichthys
11-14
11-14
12-17
8
7-10 + 6-10
34-41
8-9
2-5 + 6-14
Ceraloscppelus
13-15
13-16
12-15
8
6-7 + 6-7
35-38
9
3-5 + 9-16
Lepidophanes
11-15
13-16
11-14
8-9
6-7 + 6-8
33-37
9
3-4 + 8-1 1
Bolinichthys
11-15
11-15
11-15
8
7 + 7-8
33-36
9
3-7 + 11-17
Triphoturus
12-16
13-18
8-10
8
5-7 + 6-7
30-36
10-11
2-4 + 8-11
Slenobrachius
12-15
14-16
8-10
8
6-8 + 7-9
35-38
9-10
5-6 + 12-14
Panilux
14-17
15-18
10-13
8
8 + 8-9
35-38
10-11
4-6 + 11-15
Lampanyctus
10-19
14-21
0-17
8
6-8 + 6-8
30-40
8-11
3-8 + 9-19
• Incorrectly 15-27 in Paxlon, 1972
(1977) and Hulley (1981) on Atlantic species and McGinnis
(1982) on Southern Ocean species.
Most lantemfishes make extensive vertical migrations from
mesopelagic depths to the upper waters at night, some reaching
the surface (Paxton, 1 967). The fisheries potential of myctophids
and other mesopelagic fishes has recently been reviewed (Gjo-
saeter and Kawaguchi, 1980). Adults range in size from 20-300
mm (Kreflt, 1974) and have a life span of from one year in
some tropical species (Clarke, 1973) to more than five years in
the few temperate species that have been studied (Smoker and
Pearcy, 1970; Gjosaeter, 1973: Kawaguchi and Mauchhne, 1982).
Eggs
Myctophids are oviparous and presumably all produce plank-
tonic eggs although such have been reported for only two species.
Sanzo (1939a) indicated that mature ovarian eggs of E. rissoi
have the following characteristics: round shape; 0.80-0.84 mm
diameter; segmented yolk; single oil globule, ca. 0.28 mm di-
ameter; smooth chorion. He illustrated a planktonic egg with
similar characteristics and tentatively identified it as that o( E.
rissoi. Robertson (1977) described the planktonic egg of Lam-
panyctodes hectoris as follows: weakly oval; long axis 0.74-0.83
mm, short axis 0.65-0.72 mm; strongly segmented yolk; single
oil droplet, 0.21-0.23 mm diameter: narrow perivitelline space;
chorion smooth and delicate. He based his identification on the
similarity of these eggs and mature ovarian eggs of running ripe
L. hectoris captured at the same time by trawl.
We have observed planktonic eggs similar to those described
by Robertson (1977) but have not found them with advanced
embryos that could be matched with co-occurring yolk-sac myc-
tophid larvae. The fact that these and other types of eggs ten-
tatively identified as myctophids occur in relatively low abun-
dance compared with myctophid larvae led Moserand Ahlstrom
(1970) to suggest that the fragile chorion breaks in contact with
plankton nets and the embryo is extruded through the mesh.
Larvae
Moser and Ahlstrom (1970) reviewed the literature on myc-
tophid larvae: however, numerous recent contributions have
advanced our knowledge of the group and are listed in Table
61. Of the 32 recognized genera of myctophids, larvae have
been described for all but Hintonia. The larval stages of myc-
tophids provide sets of characters that are useful at levels of
systematic analysis from species separation to hypotheses of
222
ONTOGENY AND SYSTEM ATICS OF HSHES-AHLSTROM SYMPOSIUM
Table 61. Summary
abbreviated as follows:
OF Literature Containing Illustrations of Developmental Stages of Myctophids. Frequently cited authors are
Ahlstrom (A), Belyanina and Kovalevskaya (B + K), Dekhnik and Sinyukova (D + S), Moser and Ahlstrom (M + A),
Pertseva-Ostroumova (P-O), Shiganova (S), Tuning (T).
Species
Single larval stage
Multiple larval stages
Transforming stage
Juvenile stage
Benlhosema
fibulatum
glaciale
panamense
pterota
suborbitale
Bolinichthys
dislofax
pyrsobolus
Centrobranchus
andrae
breviroslris
choerocephalus
nigroocellatus
Ceratoscopelus
maderensis
townsendi
warming!
Diaphus
agassizii
holli
malayanus
melapoclampus
mollis
pacificus
rafinesquei
Iheta
Diogenichthys
atlanticus
laternalus
panurgiis
Electrona
antarctica
carhbergi
rissoi
subaspera
Gonichthys
coccoi
tenuiculus
Gymnoscopelus
bolini
braueri
fraseri
mcholsi
opislhoplerus
Hygophum
atraium
henoiti
brunni
hanseni
hygomi
M + A, 1974
Holt, 1898; S, 1977
M + A, 1974; P-O, 1974
P-O, 1964; M + A, 1974
M + A, 1974
P-O, 1964
P-O, 1974
P-O, 1964
M + A, 1974
P-O, 1974
M + A, 1972; S, 1977
M + A, 1974
Miller etal., 1979;
Belyanina, 1982b
D + S, 1966
M + A, 1974
P-O, 1964; M + A, 1974
P-O, 1964
M + A, 1974
M + A, 1974
M -I- A, 1974
M -I- A, 1974
P-O, 1964
P-O, 1964
P-O. 1974
T, 1918; Sparta, 1951;
M + A, 1974
M + A, 1970
Tsokur, 1981
P-O, 1974; Badcockand
Merrett, 1976; S, 1977
Holt, 1898; T, 1918
Sparta, 1951
M + A, 1970
P-O, 1974; S, 1977
P-O, 1974
P-O, 1974
M + A, 1970
-
T, 1918; D -1- S, 1966
T, 1918
S, 1977
S, 1977
P-O, 1975
T, 1918
Tsokur, 1975
Sparta, 1952
S, 1977
P-O, 1975
T, 1918
Tsokur, 1975
Sparta, 1952
S, 1977
T, 1918
T, 1918; A, 1965;
M + A, 1970; P-O, 1974;
S, 1977
A, 1965; M + A, 1970
P-O, 1974
T, 1918
T, 1918; M + A, 1970;
S, 1977
M + A, 1970
P-O, 1974
P-O, 1967; B + K, 1979 -
B + K, 1979
T, 1918; Sanzo, 1939a; Sanzo, 1939a
D + S, 1966; M -I- A, 1970
M + A. 1974
M -I- A, 1974
T, 1918; S, 1977;
D + S. 1966
M + A, 1970
S, 1977
P-O, 1977; B + K, 1979
P-O, 1977
M + A, 1972; P-O, 1977;
B + K, 1979
Yefremenko, 1977
M + A, 1970
T, 1918; S, 1974
S, 1977
T, 1918; P-O, 1974;
S, 1977
M + A, 1970
S, 1977
M + A, 1972
M + A, 1970
T, 1918; S. 1974
S, 1977
T, 1918; P-O, 1974;
S, 1977
Holt, 1898; T, 1918;
Sparta, 1951
M + A, 1970
Tsokur, 1981
S. 1977
M + A. 1970
T, 1918; S, 1977
P-O, 1975
T, 1918
Tsokur, 1975
Sparta, 1952
S, 1977
T, 1918
T. 1918; M + A, 1970;
S, 1977
M + A, 1970
T, 1918; Sanzo, 1939a;
M + A, 1970
T, 1918; S, 1977
S, 1977
M + A, 1970
T, 1918; S, 1974
S, 1977
T, 1918; S, 1977
MOSER ET AL.: MYCTOPHIDAE
Table 6 1 . Continued.
223
Species
Single larval stage
Mullipte larval stages
Transforming stage
Juvenile stage
macrochir
pro.ximum
rcinhardli
taanmgi
Idiolychnus
urolampus
Kretflichthys
anderssoni
Lampadena
luminosa
urophaos
Lainpanyctodes
hectoris
Lampanyclus
achirus
crocodilus
jordani
nohilis
pusillus
regalis
ritleri
Lampichthys
procerus
Lepidophanes
gaussi
guerjtheri
Lohianchia
M + A, 1974
M + A, 1974; Miller
et al., 1979
M + A, 1974
M + A, 1974
M + A. 1974
M + A, 1974
M + A, 1974; Miller
etal., 1979
M + A. 1974
P-O, 1964
Miller etal., 1979
M + A, 1974
M + A, 1974
M + A, 1974
M + A, 1972
S. 1975
P-O, 1974
M + A, 1970; S, 1977
Yefremenko, 1976;
B + K, 1979
M + A, 1972
Ahlstrom et al.,
1976
T, 1918; D + S, 1966
T, 1918; D + S, 1966
A, 1965
M + A, 1972
S, 1977
S, 1975
P-O, 1974
M + A, 1970; S, 1977
Yefremenko, 1976
M + A, 1972
Ahlstrom et al.,
1976
T, 1918
T, 1918
Bolin, 1939b
M + A, 1972
M + A, 1972; S, 1977
S, 1975
M + A, 1970; S, 1977
Yefremenko, 1976
Ahlstrom et al.,
1976
T, 1918
T, 1918
do/Ieini
M + A, 1974
T, 1918; D + S, 1966;
T, 1918; S, 1977
T, 1918; S, 1977
S, 1977
gemellari
Sanzo, 1931c; P-O,
1964; M + A, 1974
T, 1918
T, 1918
T, 1918
Loweina
rara
M + A. 1974
M + A, 1970; P-O, 1974
M + A, 1970
M -1- A, 1970
lerminata
Belyanina, 1982b
-
—
—
Meleleclrona
ventralis
M + A, 1974 -
—
—
—
Myctophum
asperum
P-O. 1964; M -1- A, 1974
Imai, 1958; P-O, 1974
Imai, 1958; P-O, 1974
aurolaternatum
M + A, 1974
—
__
brachygnathum
M -1- A, 1974
_
lychnobium
M + A, 1974; P-O, 1974
—
P-O, 1974
nilidulum
M + A, 1974
M + A. 1970; P-O, 1974
M + A, 1970
oblusirostre
M + A, 1974
_
—
punctalum
M + A, 1974
Sanzo, 1915b; T, 1918;
S. 1977
Sanzo, 1915b;
S, 1977
T,
1918;
T, 1918; S, 1977
selenops
M + A, 1974
—
_
spinosum
M + A, 1974
P-O, 1974
P-O, 1974
P-O, 1974
Nololychnus
valdiviae
P-O, 1964; M + A, 1974
T, 1918
T. 1918
T, 1918
Notoscopelus
caudispinosus
Belyanina, 1982b
_
—
elongatus
—
T, 1918
T, 1918
T, 1918
resplendens
M + A, 1974
M + A. 1972; Badcock and
Merrett, 1976; S, 1977
M + A, 1972;
s.
1977
224
ONTOGENfY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 61. Continued.
Species
Single larval stage
Multiple larval stages
Transforming stage
Juvenile stage
Parvilux
ingens
M + A, 1974
-
-
-
Prolomyclophum
arcticum
—
T, 1918
T, 1918
T, 1918
boHni
_
P-O, 1967; B + K, 1979
—
chilensis
M + A. 1974
—
_
—
crockeri
—
M + A, 1970
—
M -1- A, 1970
normani
P-O. 1967; M + A,
1974
_
_
P-O, 1967
parallelum
—
P-O, 1967; B -1- K, 1979
—
_
subparallelum
M + A, 1974
—
—
—
tenisoni
M + A, 1974
—
—
_
ihompsom
P-O, 1964
P-O, 1967; M + A, 1970
-
M -1- A, 1970
Scopelopsis
nndlipiinctatus
—
M + A, 1972; P-O, 1972
M + A, 1972; P-O, 1972;
M-F A, 1974
M + A, 1972
Stenobrachius
leucopsarus
P-O, 1964; M -1- A,
1974
Fast, 1960; A, 1965;
A, 1972b
Fast, 1960
Fast, 1960
Symbolophours
hoops
—
P-O, 1974
—
—
californiense
P-O, 1964; M -1- A,
1974
A, 1965; M + A, 1970;
P-O, 1974
M + A, 1970; P-O, 1974
—
evermanni
P-O, 1964
P-O, 1974
P-O, 1974
P-O, 1974
veranyi
—
Sanzo, 1915b; T, 1918;
D-l-S, 1966
Sanzo, 1 9 1 5b; T, 1918
Sanzo, 1 9 1 5b, T, 1918
Taaningichthys
minimus
-
M + A, 1972
-
-
Tarletonbeania
crenularis
P-O, 1964; M ^ A,
P-O, 1974
1974;
A, 1965; M -i- A, 1970
Bolin, 1939b; M + A, 1970
M -1- A, 1970
Tripholurus
mexicanus
M + A, 1974
A, 1965; A, 1972b
_
_
nigrescens
Moser, 1981
-
-
-
Op
po XXvo
Fig. 1 14. Hypothetical myctophid showing photophore terminology, from Paxton (1972).
MOSER ET AL.: MYCTOPHIDAE
225
Fig. 1 15. Larvae of Electronini. (A) Krefftichlhys anderssoni. 15.7 mm; (B) Protomyctophum normani. 15.2 mm; (C) P. Heirops ihompsom,
13.8 mm; (D) Elcclrona rissoi. 7.9 mm; (E) £. antarclica. 12.7 mm; (F) Melelectrona ventralis, 10.3 mm. A, B, E, F from Moser and Ahlstrom
(1974); C and D from Moser and Ahlstrom (1970).
226
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 62. Sequence of Formation of Photophores which Appear in Fourteen Genera of Myctophidae. The Bfj appear first in all genera
listed. Parentheses indicate photophores appear late in larval period.
PO,
PO,
PVO, PLO
VO, AOa, AOa,
Benthosema
suborbitale
glaciate
pterola
fibulalum
Diogenichthys
lalernalus
atlanticus
Myclophum
spinosum
lychiwbium
asperum
brachygnalhum
obtusirosire
selenops
Lobianchia
Diaphus
theta
pacificus
Gymnoscopelus
Lampanyctodes
Scopelopsis
Lainpichthys
Notoscopelus
Lampadena
Ceratoscopelus
Lepidophanes
Bolinichlhys
22--2 1 1 333------33
- - - - (1) (1) (1) (1) (1) (1) ------- -
--1-4 6---2 3 5--5-6-
--1--3 5--2---6--46
_________ 1 ________
(5)
1
2
2
2
3
1
1
(1)
(4)
(3)
(7) (8)
(5) -
(4)
(1)
- (9) (3) (6)
- - (6) -
3
3
1
2
1
(1)
ordinal relationships. One set is the size at various develop-
mental milestones. Myctophid larvae hatch at about 2 mm length
with a yolk-sac remnant. Notochord flexion occurs in a narrow
size interval (0.5-2.0 mm) and the size at mid-flexion is typically
about half the maximum larval size. Size at transformation also
occurs within a short length interval, usually not exceeding 2
mm. Most myctophid species transform in the length range of
12-19 mm, although some (e.g., Electrona rissoi, Notolychnus
valdiviae) are as small as 9-10 mm at transformation and some
species of Symbolophorus reach about 23 mm before transfor-
mation. Gymnoscopelus nicholsi has the largest larvae recorded,
up to 28 mm.
Head, body, and gut shape are distinctive for most species
and within most genera there is a similarity of shape (Figs. 1 1 5-
124). While most myctophid larvae are moderately slender,
body shape can range from highly attenuate (e.g., Hygophum
reinhardti) to markedly robust (e.g., some Myctophum and
Lampanyctus species). Some are deep-bodied but laterally com-
pressed (e.g., Gonichthyini). Robust larvae and deep-bodied,
laterally compressed forms tend to have large heads and jaws,
while attenuate forms have flat heads.
The eye is varied in size and shape and provides numerous
characters. In the Myctophinae the eyes are elliptical in outline
in contrast to most Lampanyctinae which have rounded eyes.
Further specializations in Myctophinae are the presence of var-
iously shaped choroid tissue on the ventral surface of the eye
in most genera and eye stalks in several genera. Among 1am-
panyctine genera eyes are sessile and only Lobianchia doflcini
and species of Triphoturus have markedly narrowed eyes with
choroid tissue.
The gut has distinctive transverse rugae and ranges from short,
to elongate, to trailing free from the body. In most myctophids
it extends to about the midpoint of the body and is slightly S-
shaped. The curvature tends to be more pronounced in taxa
with short guts. In two myctophine genera (Metelectrona and
some Hygophum species) the anterior section of the gut is small
in diameter and opens dorsally into the relatively larger pos-
terior section.
In most myctophids, ray formation and ossification of fins
proceeds in the following sequence: caudal, pectoral, anal, dor-
sal, and pelvic. However, in some Symbolophorus species the
pelvic fin forms early and ossification of rays precedes that of
the anal and dorsal fins. In most species the pectoral fin is
relatively small, but deep-bodied and robust forms in both
Fig. 1 16. Larvae of Myctophini. (A) Benthosema glaciale. 10.5 mm; (B) B. suborbitale. 9.2 mm; (C) B. pterola. 8.5 mm; (D) B. fibulatum.
8.7 mm; (E) Diogenichthys lalernalus. 1.1 mm; (F) D. atlanticus. 8.8 mm. A-D from Moser and Ahlstrom (1974); E and F from Moser and
Ahlstrom(1970).
^^JmiJdJ^i^L
*ss;;;^a
228
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 1 17. Larvae of Myctophini. (A) Hygophum proximum. 8.9 mm; (B) H. taaningi. 6.8 mm; (C) H. reinhardti. 12.8 mm; (D) Symbolophorus
californiense. 1 1.5 mm; (E) Myctophum punclalum. 13.6 mm; (F) M. aurolalernatum. 26.0 mm. A. B. E, F from Moser and Ahlstrom (1974); C
and D from Moser and Ahlstrom (1970).
MOSER ET AL.: MYCTOPHIDAE
229
subfamilies have large fins and fin bases. In Symbolophorus the
fin base is uniquely shaped and in Lobianchia the fin blade has
a unique shape. In two genera (Loweina. Tarletonbeania) the
lowermost pectoral ray is elongate and ornamented. The finfold
is enlarged in many myctophine genera and greatly enlarged in
one myctophine tribe, the Gonichthyini.
Myctophids, with the exception of Notolychnus and Taan-
ingichthys, develop the middle branch iostegal photophore (Br,)
during the larval period. It is located posteroventral to the orbit
but during transformation assumes a position beneath the orbit
on the branchiostegal membrane. Three myctophine genera and
1 1 lampanyctine genera develop additional photophores during
the larval period; however, the Br, is always the first to develop.
The larval photophore complements and the sequence of ap-
pearance of constituent photophores are useful characters.
Myctophid species have distinct melanophore patterns, with
the exception of the large genus Diaphus, for which only a few
specific patterns have been identified. Most genera may be sep-
arated by overall similarity of pattern among their species and
some have unique melanophore loci. There are no clear patterns
for tribes or subfamilies although certain pigment loci are per-
sistent in some tribes (e.g., caudal fin base spots in diaphines;
dorsal midline series in gymnoscopelines).
In the following summary of key larval characters, the genera
are listed for convenience as in Moser and Ahlstrom (1970.
1972, 1974) and the sequence does not necessarily imply rela-
tionship. Likewise, the species groups serve only to identify
phenotypically similar larval types. Larvae of a majority of myc-
tophid genera have a moderately slender body, a head of mod-
erate size, with a slightly convex dorsal profile and a pointed
snout of moderate length. Body and head shape are noted only
when they depart from this morph. In Myctophinae eye shape
is noted when it is markedly elliptical and size is noted only
when larger or smaller than typical. In Lampanyctinae eye shape
is noted only when it departs from the round condition and eye
size only when larger or smaller than typical. Choroid tissue is
described only when it is present. Gut length and shape are
described only if there is a departure from the typical morph —
a slightly S-shaped gut that extends to about midbody. The most
persistent pigment locus in myctophid larvae is above or to the
side of the free terminal section of the gut, thus only the lack
of this pigment is noted. Larval photophores, in addition to the
Br,, and their sequence of appearance are shown in Table 62.
Myctophinae
Krefflichthys. — Fig. 1 15 A; head small with short snout; conical
choroid tissue; gut straight, extending beyond midbody; dorsal
fin displaced posteriad; lateral gut and postanal median ventral
melanophore series; large lateral hypural pigment patch.
Protomyctophum. — ¥\%. 1 15B, C; two subgenera; head small to
moderate in size; gut short, wide space between anus and anal
fin; head pigment lacking except in otic region of P. Heirops
chilensis; some species may have melanophores on lateral gut,
above gut on trunk, above gas bladder, in postanal ventral mid-
line series, prominent pigment on lateral hypural region. P.
Heirops: Fig. 1 1 5C; characters similar to P. Protomyctophum
except eye narrower.
Eleclrona— Fig. 1 15D, E; body moderately slender to moder-
atey deep; head moderately large; snout blunt or pointed; gut
short, somewhat saccular, strongly S-shaped; space between anus
and anal fin not as large as in Protomyctophum; three morphs.
E. subaspera-E. carlsbergi: eye slightly elliptical, small lunate
choroid mass in E. carlsbergi; pigment above gut; E. subaspera
has pigment lateral to cleithrum. E. rissoi: Fig. 1 1 5D; head large,
broad; eye very narrow; pigment at lower jaw symphysis, on
pectoral fin blade. E. antarctica: Fig. 1 1 5E; body and head lat-
erally compressed; gut mass protrudes ventrally from body pro-
file; eye small, narrow, with bicolored elongate conical choroid
mass; pigment on upper jaw, pectoral fin blade, lateral gut,
lateral hypural region.
Metelectrona. — Fig. 1 1 5F; body and head laterally compressed;
dorsal finfold enlarged with fin base initially separated from
body; lunate choroid mass; anterior gut section with small di-
ameter, opening dorsally into somewhat saccular posterior sec-
tion; pigment below lower jaw and on isthmus.
Benthosema. — Fig. 1 16A-D; two morphs; photophores (Table
62). B. glaciale-B. sitborbitale: Fig. 1 16A, B; eyes narrow, with
small lunate choroid mass; gut moderately short in preflexion
larvae with space between anus and anal fin; pigment on snout,
lower jaw, hindbrain, lateral and ventral cleithral region; pig-
ment above gut in B. glaciate. B. pterota-B. fibulatum: Fig. 1 I6C.
D; eyes less narrow than in above morph, with sliver of choroid
tissue or none; gut extends to about midbody with no space
between anus and anal fin; preflexion larvae with melanophore
series on lateral gut and on postanal ventral midline, coalescing
to a single melanophore; lateral cleithral pigment; lower jaw
pigment in B. pterota.
Diogenichthys. — Fig. I16E, F; eyes very narrow in preflexion
stage, less so in postflexion; photophores (Table 62); pigment
series on lateral gut and on postanal ventral midline, increasing
with development; spot at caudal fin base; pigment on tip of
lower jaw in D. laternatus; D. atlanticus has spot on trunk above
terminal gut flexure and pigment on symphyseal barbel.
Fig. 1 18. Larvae of Myctophum. (A) M. phengodes. 9.8 mm; (B) M. asperum. 6.8 mm; (C) M. brachygnathum. 7.5 mm; (D) M. selenops. 7.8
mm; (E) A/, spinosum, 9.0 mm. From Moser and Ahlstrom (1974).
Fig. 1 19. Larvae of Gonichthyini. (A) Loweina rara. 17.6 mm; (B) Tarletonbeania crenularis. 18.9 mm; (C) Gomchthys tenutculus. 1.1 mm;
(D) Centrobranchus choerocephalus. 7.3 mm. From Moser and Ahlstrom (1970).
Fig. 120. Larvae of Lampanyctinae. (A) Notolychnus valdiviae. 8.7 mm; (B) Lobianchia dojleini. 8.2 mm; (C) L. gemellari. 6.7 mm; (D)
Diaphus theta. 6.9 mm; (E) D. pacificus. 5.2 mm; (F) Gymnoscopelus nicholsi. 23.5 mm. A-E from Moser and Ahlstrom (1974); F from Moser
and Ahlstrom (1972).
Fig. 121. Larvae of Lampanyctinae. (A) Lampanyctodes hectoris. 1 3.0 mm; (B) Scopelopsis muttipunctatus. 1 3.4 mm; (C) Lampichthys procerus,
1 4.5 mm; (D) Notoscopelus resplendens. 11.2 mm; (E) Lampadena lununosa. 1 2.8 mm; (F) Taanmgichthys minimus. 1 4.4 mm. A from Ahlstrom
et al. (1976); B, C. F from Moser and Ahlstrom (1972); D and E from Moser and Ahlstrom (1974).
230
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
^yig^i®^'
^^^>
MOSER ET AL.: MYCTOPHIDAE
231
232
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
MOSER ET AL.: MYCTOPHIDAE
233
234
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Hygophum. — Fig. 1 17A-C; diagnostic pattern of melanophores
at the cleithral symphysis and isthmus region consisting of paired
pigment dashes that form a median Hne as the series extends
forward on the isthmus; Br, photophore forms late in larval
period; three morphs. H. proximum-H. hygomi-H. benoiti-H.
hanseni-H. brunni: Fig. 1 1 7A; eye moderately narrow with con-
ical choroid tissue; pigment sparse in most species with some
lateral gut spots in all species; some species may have pigment
on hypaxial myosepta, jaws, lateral cleithral region, base of cau-
dal rays. H. atratuin-H. reinhardti: Fig. 1 1 7C; body very slen-
der; head flat; eyes very narrow, on short stalks; elongate conical
choroid mass; gut almost straight, small diameter; pigment se-
ries along lateral gut and hypaxial myosepta; pigment at caudal
fin base; pigment on lower jaw symphysis in H. atratum. H.
macrochir-H. taaningi: Fig. 117B; body and head deep and
laterally compressed; eyes large, relatively wide; no choroid tis-
sue; anterior gut section narrow in diameter, opening dorsally
into somewhat saccular posterior section; H. macrochir has pig-
ment on upper and lower jaw and a patch of melanophores on
posterior gut section; H. taaningi has pigment on gular region
and lateral surface of cleithrum.
Symbolophorus. — Fig. 1 17D; head broad, somewhat flat; eyes
slightly stalked, conical choroid mass; pectoral fin large with
supernumerary rays, base wing-shaped, rays ossify early; pelvic
fin large, early-forming in some species; dorsal finfold well de-
veloped with fin base forming in it; pigment series on lateral
gut and postanal ventral midline in preflexion larvae; pigment
on snout, hindbrain, lateral cleithral region, isthmus, paired fins.
Myctophum.— Figs. 1 17E, F and 1 18A-E; at least five distinct
morphs, all but M. aurolaternatum with enlarged fan-shaped
pectoral fins, some with supernumerary rays and early ossifi-
cation; conical choroid mass. M. aurolaternatum: Fig. 117F;
body very slender; head somewhat flat; eyes small, on elongate
stalks; gut straight, at midbody becomes trailing, extending to
well beyond caudal fin; dorsal finfold well developed, fin base
forms at its margin; pigment series on lateral gut, evenly dis-
tributed on trailing section, except heavier near terminus; pig-
ment on jaws, isthmus, opercle, branchiostegal membrane, pec-
toral fin, anal fin base, caudal fin. M. nitidulum-M . punctatum:
Fig. 1 1 7E; body moderately slender to slightly deep; head broad,
somewhat flat in preflexion stage; eyes on short stalks; numerous
small melanophores on snout, jaws, brain, isthmus, branchio-
stegal membrane; two rows of melanophores on ventral surface
of gut; opposing melanophores on postanal dorsal and ventral
midline; pigment on pectoral fin base and blade and at base of
caudal rays. M. phengodes: Fig. 1 1 8 A; body and head moder-
ately deep; similar to M. nitidulum, except pigment sparse and
eyes not stalked; pigment at base of pectoral fin rays. M. spi-
nosum-M. lychnohium: Fig. 1 18E; head with convex dorsal pro-
file and long snout giving the larva a fusiform appearance; long
axis of eye rotated towards horizontal; photophores (Table 62);
head heavily pigmented on jaws, brain, postorbital and oper-
cular regions; pigment above gut on trunk, embedded in my-
osepta in M. spinosum; opposing dorsal and ventral midline
blotches, larger and more deeply embedded in M. spinosum
with embedded myoseptal pigment along horizontal septum;
blotch at base of caudal rays. M. asperum-M . brachygnathum-
M. obtusirostre-M. selenops: Fig. 118B-D; body deep, robust;
head broad, deep with convex dorsal profile and large snout;
eye relatively larger than in other morphs; choroid tissue broadly
conical, except in M. selenops where it is elongate and pigmented
at tip; photophores (Table 62); head pigment similar to M.
spinosum; most species have heavy pigment lateral to cleithra
and on pectoral fin bases; all species lack trunk and tail pigment,
except M. asperum which has extensive embedded myoseptal
and dorsal/ventral midline blotches.
Loweina. — Fig. 1 1 9 A; body and head moderately deep, laterally
compressed; dorsal and anal fins displaced far posteriad; dorsal
and ventral finfolds greatly enlarged and conspicuously pig-
mented to produce a disc-shaped profile; eyes large; gut with
expanded anterior section and enlarged terminal section; pec-
toral fin large with lower-most ray elongate, ornamented with
pigmented spatulations; interorbital pigment band; pigment at
lateral cleithral surface, dorsal fin origin, and opposing midline
blotches at caudal peduncle region.
Tarletonbeania. — Fig. 1 198; similar to Loweina. except median
fins displaced less posteriad; eye narrower and with lunate cho-
roid mass; four melanophores on periphery of brain, two me-
lanophore series on ventrum of gut.
Gonichthys. — Fig. 1 1 9C; body and head deep and laterally com-
pressed, leaf-like; snout large, angulate in profile; eye small with
elongate conical choroid mass, pigmented at tip; enlarged dorsal
and ventral finfolds; pectoral fins moderately large; pigment on
snout, jaws, midline of brain, postorbital and opercular regions;
pigment on lateral hindgut and on trunk above gut; series of
embedded blotches on dorsal midline of body, opposing blotch-
es on postanal ventral midline; large pigment patch on lateral
caudal peduncle region in G. tenuiculus; heavy embedded pig-
ment streak along horizontal septum in G. coccoi.
Centrobranchus. — Fig. 119D; morphology similar to Gonich-
thys except snout markedly blunt and rounded and terminal gut
flexure less acute; two morphs. C. choerocephalus-C. breviros-
tris-C. nigroocellatus: Fig. 1 19D; eye very narrow with unpig-
mented choroid mass that exceeds it in length; pigment sparse;
some at postorbital-opercular region, branchiostegal membrane,
ventral surface of liver. C. andrae. eye wider than in above
morph and with short conical choroid mass; pigment extensive,
on snout, upper jaw, dorsal brain, opercle, branchiostegal mem-
brane, lateral hindgut, ventral surface of liver, pectoral fin base;
embedded spots along dorsal midline with opposing spots along
postanal ventral midline; embedded spots along horizontal sep-
tum in caudal peduncle region.
Lampanyctinae
Notolychnus. — Fig. 1 20A; head relatively large with moderately
elongate snout; eyes usually narrow, often irregular in shape; gut
short, more so in preflexion stage; no photophores, even Br,
lacking; pigment on lateral hindgut, gas bladder, base of caudal
rays; a persistent but sparse postanal ventral midline series.
Lobianchia. — Fig. 120B, C; body deep, robust; head broad with
large snout; pectoral fins large; blade wing-shaped with upper
rays longer than others; photophores (Table 62); head unpig-
mented; pigment on trunk, on gut below pectoral fin base, on
pectoral fin base and blade, embedded in gut region anterior to
pectoral fin base, along anal fin base, and at base of caudal rays;
embedded melanophores in myosepta above pectoral fin be-
coming extensive in postflexion stage; two morphs. L. dofleini:
Fig. 120B; eye small, narrow, with lunate to squarish choroid
MOSER ET AL.: MYCTOPHIDAE
235
Fig. 122. Urvae of Lampanyctmae. (A) Ceraloscopehis townsendi. 16.6 mm; (B) Lepidophanes gaussi. 13.5 mm; (C) BoUmchthvs distofax.
9.4 mm; (D) Slenohrachius leucopsarus. 10.4 mm; (E) Parvilux ingens. 14.4 mm; (F) Triphoturus mexicanus. 10.5 mm. A-E from Moser and
Ahlstrom (1974); F from Ahlstrom (1972b).
236
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
mass; gradual transition from lower pectoral rays to longer upper
rays. L. gemellari: Fig. 120C; eye large, almost round, choroid
mass a lunate sliver; abrupt transition between lower pectoral
rays and long upper rays.
Diaphus. — Fig. 1 20D, E; pigment lacking on head; melanophore
at anteroventral surface of liver, one or more at midgut region,
one or more at base of caudal rays; gas bladder pigmented; two
morphs. D. theta: Fig. 120D; body moderately slender; head
moderate in size; photophores (Table 62); numerous melano-
phores in postanal ventral midline series, persisting into post-
flexion stage. D. pacificus: Fig. 120E; body moderately deep,
somewhat robust; head moderately large; photophores (Table
62); a few melanophores in postanal ventral midline series, usu-
ally coalescing to one before flexion stage.
Gymnoscopelus. — Fig. 120F; photophores (Table 62); pigment
above brain, at lateral cleithral region, above midgut, above gas
bladder; postanal ventral midline series present but, in some
species, restricted to caudal peduncle region; melanophore series
on each side of dorsal midline, in most species extending be-
tween caudal and dorsal fins, in others extending forward to
dorsal fin origin, and in others restricted to caudal penduncle
region; pigment at base of caudal rays; some species have pig-
ment on lateral hypural region; lateral pigment patch at caudal
peduncle in G. opisthopterus, which also has embedded mela-
nophores above vertebral column.
Lampanyctodes. — Fig. 121 A; photophores (Table 62); pigment
above brain, at anteroventral surface of liver, above gas bladder;
a postanal ventral midline series and a series on each side of
dorsal midlme between dorsal and caudal fins; pigment at base
of caudal rays and at lateral hypural region.
Scopelopsis. — Fig. 121B; photophores (Table 62); pigment sim-
ilar to Lampanyctodes except additional melanophores on hind-
brain, nape, lateral cleithral region; pigment rows along dorsum
irregular.
Lampichthys. — Fig. 121C; photophores (Table 62); pigment
similar to Scopelopsis except dorsal rows consist of large closely-
spaced melanophores which at maximal development extend
from caudal fin to dorsal fin origin; a short melanophore series
along horizontal septum on caudal peduncle in late postflexion
stage.
Notoscopelus. — Fig. 12 ID; photophores (Table 62); body mod-
erately deep; head moderately large; eye large; snout becomes
somewhat bulbous at flexion stage; gut short in early preflexion
stage, elongates to about midbody by late preflexion; pigment
at tips of jaws, above brain, above gas bladder and at lateral
cleithral region in early postflexion larvae; additional pigment
develops below lower jaw, on hindbrain and nape; series of
melanophores on each side of dorsal midline, beginning at mid-
body and gradually developing along entire dorsum; series along
horizontal septum and along anal fin base; pigment on base of
caudal rays and on pelvic and anal rays in some species at late
postflexion stage; extensive embedded myoseptal pigment on
trunk or tail in postflexion stages of some species.
Lampadena. — Fig. 1 2 1 E; photophores (Table 62); pigment above
brain, nape, gut, gas bladder; most species have large melano-
phores along dorsal midline, with opposing postanal ventral
midline melanophores; some species with smaller, more nu-
merous melanophores in dorsal and ventral series; embedded
pigment above spinal column in some species.
Taaningichthys. — Fig. 121F; body slender; lower jaw projects
beyond upper; no photophores, even Br, lacking; pigment above
brain, in otic region, one to several opposing melanophores at
postanal dorsal and ventral midline; late postflexion larvae may
develop minute melanophores along each side of dorsal midline;
pigment at base of caudal rays; series of embedded melano-
phores above spinal column.
Ceratoscopelus. — Fig. 122A; eye elliptical in early larvae; pho-
tophores (Table 62); pigment above gut; postanal ventral mid-
line series in early larvae, coalesces to a single spot in postflexion
larvae; C. maderensis has short series at dorsal and ventral
midline in caudal peduncle region; embedded pigment above
posterior region of spinal column in some species.
Lepidophanes. — Fig. 122B; eye small; photophores (Table 62);
usually two melanophore pairs at dorsal midline in caudal pe-
duncle region and one or two ventral midline melanophores; L.
gaiissi has median melanophore above hindbrain and median
ventral melanophore below pectoral fin base.
Bolinichthys. — Fig. 1 22C; moderately deep-bodied; snout blunt;
eye large; photophores (Table 62); sparse pigment; midline spot
above brain, embedded otic spot, embedded pigment above gut;
some species with a sparse postanal median ventral series that
coalesces to a single melanophore; B. distofa.x has a short series
on horizontal septum; embedded pigment above posterior re-
gion of spinal column in some species.
Triphoturus. — Fig. 122F; eye elliptical with choroid mass; pig-
ment at tip of lower jaw, at angular region of jaw, at lateral
cleithral region; early preflexion larvae have paired lateral gut
spots near pectoral fin base and at midgut; anterior pair coalesces
to a median position anteroventral to liver, the posterior pair
becomes dorsal to gut; pigment above gas bladder; early pre-
flexion larvae have postanal median ventral series that coalesces
to one or two spots; pigment along margin of preanal finfolds;
a single dorsal spot at adipose fin in T. mexicanus; a series of
pigment dashes on horizontal septum in T. nigrescens.
Stenobrachius. — Fig. 1 22D; gut melanophores and postanal me-
dian ventral series similar to Triphoturus; pigment above brain
and nape in postflexion stage; late postflexion larvae have
embedded melanophores in trunk myosepta and melanophore
series on each side of dorsal midline.
Parviln.x. — Fig. 122E; head, eyes large; tapered body; gut short
Fig. 123. Larvae oi Lampanyclus. (A) L. steinbecki. 6.6 mm; CalCOH Sla. 70.200; (B) L. pusiUus. 1.1 mm; redrawn from Taaning (1918);
(C) L. nobilis, 9.6 mm; SEFC, OR II 7343 Sta. 98; (D) L. par\icauda. 7.5 mm. SWFC, Eastropac Op Sta. 023; (E) L. crocodilus. 11.5 mm, redrawn
from Tining (1918).
MOSER ET AL.: MYCTOPHIDAE
237
Fig. 124. Larvae of Lampanyclus. (A) L. rilleri. 10.1 mm; (B) L. idostigma. 7.2 mm. CalCOFI 6002 Sta. 133.45; (C) L. regalis. 13.0 mm;
(D) Lampanyctus sp., 8.7 mm; (E) L. achirus. 13.4 mm; (F^ Lampanyclus sp., 9.4 mm. A, C, D, E from Moser and Ahlstrom (1974); F from
Moser(1981).
MOSER ET AL.: MYCTOPHIDAE
239
in early preflexion stage, elongates to midbody by flexion stage;
in postflexion stage pigment above brain, embedded in otic
region, lateral to cleithrum, at anteroventral region of liver; one
to several dorsal median melanophores and one ventral median
melanophore at caudal peduncle.
Lampanyclus. — Figs. 123. 124; body slender; head deep; gut
short in early preflexion stage; during preflexion stage gut length-
ens to midbody. body deepens and becomes somewhat robust
in most species; pigment above brain in most species; postflex-
ion larvae develop trunk myoseptal pigment that increases to
cover most of the anterior trunk at transformation; at least 6
morphs. L. nohilis-L. parvicaiida-L. oinostigma-L, crocodilus-
L. ritteh-L. idostigma: Figs. 123C-E. 124A, B; body and head
moderately deep; eyes, jaws, pectoral fins moderate in size; pig-
ment may be present at snout, lower jaw. opercle, above gut,
anteroventral surface of liver, at dorsal or ventral midline on
tail. L. pusillus-L. steinbecki: Fig. 123 A, B; deep, broad body
and head, very robust; snout blunt; eyes large; dorsal and anal
fins displaced posteriad; pectoral fins moderately large; L. pus-
illus heavily pigmented on head, body, pectoral fin base; series
along horizontal septum; L. steinbecki with pigment below lower
jaw, on opercle. pectoral fin base; series along horizontal septum
and embedded pigment on tail in postflexion larvae. L. regalis-
L. ater. Fig. 1 24C; deep, large head and body; snout elongate,
jaws large, teeth well developed, especially at tip of upper jaw;
preopercular spines in some species; dorsal and anal fins dis-
placed posteriad; pectoral fins moderate to large; pigment may
be present at tips of jaws, embedded in snout, at postorbital
and opercular regions, pectoral and pelvic fins; spot at adipose
fin in L. regalis; one or two dorsal spots in L. ater. Information
on L. ater irom H. Zadoretsky (Dept. Zoology, Univ. of Rhode
Island, pers. comm.). L. achirus: Fig. 1 24E; body moderately
deep; head and jaws large with snout produced into toothy ros-
trum; dorsal and anal fins displaced posteriad; pectoral fins mod-
erately large; pigment on tips of jaws, embedded in snout, and
present at postorbital and opercular regions. L. lineatus-L. cu-
prarius: body moderately elongate; snout elongate, jaws large;
head pigment as in L. achirus; L. lineatus pigment consists of
numerous melanophores along dorsum and ventrum and at base
of caudal rays; L. cuprarius has pigment above gut and an ir-
regular bar below dorsal fin. Information from H. Zadoretsky
(pers. comm.).
(H.G.M.) National Marine Fisheries Service, Southwest
Fisheries Center, P.O. Box 271, La Jolla, California
92038; (J.R.P.) The Australian Museum, 6-8 College
Street, Sydney 2000, Australia.
Myctophidae: Relationships
J. R. Paxton, E. H. Ahlstrom and H. G. Moser
THE family Myctophidae has usually been placed in the order
Myctophiformes (Iniomi. Scopeliformes) since the work
of Regan (191 la), who recognized two suborders, the Mycto-
phoidea and Alepisauroidea (ateleopodids, given a third sub-
order, are currently placed elsewhere). The families Myctophi-
dae and Neoscopelidae have long been considered close relatives;
they were placed in one family until 1949 (Smith). Although
Greenwood et al. (1966:371) relegated the order to a subordinal
level within the Salmoniformes, they pointed out that mycto-
phoids. and neoscopelids in particular, possess advanced char-
acters that indicate they may be ancestral to the paracanthop-
terygian radiation. Paxton (1972:54-55)considered myctophids
and neoscopelids most closely related to the Chlorophthalmi-
dae. with that evolutionary line of the Myctophoidea arising
from an aulopid-like ancestor. Moser and Ahlstrom (1970: 141-
142) described the larval similarities in the families Chloroph-
thalmidae, Neoscopelidae and Myctophidae.
Family Relationships
Rosen (1973, 1982) split ofl" the Myctophidae and Neosco-
pelidae as a restricted order Myctophiformes which he consid-
ered the primitive sister group of both the Paracanthopterygii
and Acanthopterygii; the remaining myctophiform families were
placed in a new order Aulopiformes. Matsuoka and Iwai (1983)
found cartilage in the adipose fin of only the Myctophidae and
Neoscopelidae in the five 'iniomous' families they studied. Oki-
yama (1974b) studied the relationships of the suborder Mycto-
phoidea (sensu Gosline et al., 1966) and based on larval peri-
toneal pigment spots and the relationship of abdominal to caudal
vertebrae, three familial groups were recognized: Aulopidae-
Synodontidae-Bathysauridae, Chlorophthalmidae-Ipnopidae
and Neoscopelidae-Myctophidae. Sulak (1977) lumped the
Ipnopidae and Bathypteroidae into the Chlorophthalmidae and
the Harpadontidae and Bathysauridae into the Synodontidae,
considering both groups arose from the Aulopidae; he did not
consider the position of the Myctophidae. Schwarzhans (1978)
considered myctophids and neoscopelids most closely related
and distinct from Aulopiformes on the basis of otolith mor-
phology.
In his excellent study of the Evermannellidae. Johnson ( 1 982)
presented a rigorous analysis of 5 1 characters involving mostly
adult but some larval features. He concluded that neoscopelids
and myctophids are most closely related to each other, sharing
eight derived character slates, but that they were the sister group
of four families (Notosudidae, Scopelarchidae, Chlorophthal-
midae and Ipnopidae) constituting a chlorophthalmoid group
within the Myctophiformes. However, he noted only a single
shared derived character in those six families, and it is shared
with part of another line. Johnson (1982:95) placed the Aulo-
pidae in a second line and all remaining families in the third
240
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 63. Characters of the Myctophidae. (0) = plesiomorphic
state, (1) = apomorphic state, (2) = different or advanced apomorphic
state, 1 = by outgroup comparison, 2 = raised photophore, 3 = gener-
alized larva, * = discussed in text.
Characters
1. Jaws long (0). moderate (1), short (2)—*.
2. Extrascapulars 2 (0), 1 from fusion (1), 1 from loss (2) — *.
3. Cleithral shelf absent (0), present (1)— 1.
4. Pre 3-9 (0?), 1-2(1?)-*.
5. Larval eyes round (0), narrow ( 1 )— 1 , 3.
6. Dn present (0?), absent (1?)-*.
7. Moderately or strongly hooked teeth in posterior dentary absent
(0), present (1)-1.
8. Procurrent ventral rays 5-10 (0), 9-15 (1)-1.
9. Supramaxillary present (0), absent (1)— 1, *.
10. PO4 level (0), raised (l)-2.
1 1 . Pubic plate narrow (0), wide ( 1 )— 1 .
12. PO, and PO, level (0), raised (l)-2.
13. VO, level (0), raised (l)-2.
14. PVO horizontal (0). angled (1), vertical (2)-2.
15. Caudal luminous organs present (0), absent (1)—*.
16. AOa, level (0), raised (l)-2.
17. Pol angled (0), horizontal (l)-2, *.
18. Enlarged teeth in dentary absent (0), present (I)— I.
19. Vertebrae 28-41 (0), 41-45, (1)-1, *.
20. VO, level (0), elevated (1)- 2.
2 1 . Enlarged dentigerous area on anterior premaxillary absent (0), pres-
ent (1)-1.
22. Secondary photophores absent (0). present ( 1 )— 1 .
23. Larval gut moderate (0), initially short (1), long (2) — 3, *.
24. Larval trunk myoseptal pigment absent (0), present (1)— 1, 3.
25. Slightly hooked teeth in posterior dentary absent (0), present ( 1 )—
I.
26. Caudal luminous organs not sexually dimorphic (0), sexually di-
morphic (1)—*.
27. Larval photophores (except Br,) absent (0), present (1)— 1, 3. *.
28. Hyomandibular foramen behind anterior head (0), in anterior head
(1)-1.
29. Accessory luminous tissue absent (0), present (1)— 1.
30. Caudal luminous organs any other state (0), homogeneous and
translucent ( 1)—*.
3 1 . Procurrent ventral rays without hooks (0), with hooks ( 1 )— 1 .
32. Procurrent dorsal rays without hooks (0), with hooks (1)— 1.
33. Crescent of white tissue on posterior iris absent (0), present (1) —
1.
34. Pol 0(0), 1 (1), 2-3 (2)- 2, *.
35. Dorsal process of opercular head of hyomandibula absent (0), pres-
ent (1)-1.
36. SAOs weakly angled (0), strongly angled (I) — 2, *.
37. Larval eyes moderate (0), very large (1)— I, 3.
38. PLO level with PVO, (0), above PVO, (l)-2.
39. SAO 2, close to VO and AO series (0), 2-3 above VO and AO
series (1)— 2.
40. Larval pectoral fin moderate (0), large (1)— 3, *.
41. Mouth terminal (0), subtcrminal (1)— 1.
42. Antorbital broad (0), thin (1)— 1.
43. Larval fin fold small (0), extensive (1)— 1, 3.
44. PLO below (0) opposite or proximate to upper pectoral base (1),
far above upper pectoral base (2)— 2.
45. Lower pharyngeal teeth conical (0), pegs or plates (1)— 1.
46. Nasal trough-shaped (0), convex (1)— 1.
47. Larval lower pectoral ray not elongate (0), elongate (1)— 1, 3.
48. Gill rakers lathe-like (0), as tooth plates (1)— 1.
49. Dorsal hypurals 4 (0), 3-2 (1). 1 (2)- I.
50. Coracoid fenestra present (0), absent (1)— 1.
51. Double row of isthmus pigment in larvae absent (0), present (1) —
1, 3.
52. Premaxillary teeth conical (0), flattened (1)— 1.
53. Larval pectoral base fan-shaped (0), wing shaped (1)— 1, 3.
54. Larval head pigment present (0), absent (1)— 1, 3.
Table 63. Continued.
55. Larval choroid tissue absent (0). present (1)— 1,
56. Larval body width moderate (0), thin (1)— 1, 3.
57. Larval gut uniform (0), bipartite ( 1 )— 1 , 3.
58. Ossified distal pectoral radials 0 (0), 1-7 (1)— 1.
59. CO, keel or ridge absent (0), present (1)— 1, *.
group (the alepisauroids plus synodontoids) in his arrangement
of the order. We do not have further evidence to present in
favour of any of the above hypotheses (but do note the coiled
gut of neoscopelid lai-vae resembles the condition found in higher
groups).
Generic Relationships
Paxton (1972) analyzed features of the osteology and pho-
tophore patterns of the Myctophidae and presented a taxonomy
outlining his views of evolutionary relationships that included
two subfamilies (Myctophinae and Lampanyctinae), six tribes
(Myctophini, Gonichthyini, Notolychnini, Lampanyctini, Dia-
phini and Gymnoscopelini), 28 genera and two subgenera. The
Myctophinae was considered the more primitive of the subfam-
ilies, while the monotypic Notolychnini was provisionally placed
in the Lampanyctinae. In four papers Moser and Ahlstrom ( 1 970,
1972, 1974; Ahlstrom et al., 1976) detailed the larval charac-
teristics of all but two genera of Myctophidae and translated
their findings into a picture of evolutionary relationships. The
relationships proposed by Paxton and Moser and Ahlstrom were
strikingly similar overall and in many details. The larval studies
supported the recognition of two subfamilies composed of the
same genera indicated by the adult analysis, highlighted the
enigmatic features of Notolychnus. and recognized three addi-
tional tribes in the Lampanyctinae. Notable differences in the
conclusions of the two studies included consideration of the
Lampanyctinae as the most primitive subfamily by Moser and
Ahlstrom, non-recognition of the tribe Gonichthyini ( Tarleton-
beama. Loweina. Gonichthys, Ccntrohranchus) as a monophy-
letic taxon in the larval study, inclusion of the genera Taan-
ingichthys. Lampadena. Bolinchthys, Lepidophanes and
Ceratoscopelus in the tribe Gymnoscopelini by Moser and Ahl-
strom and the tribe Lampanyctini by Paxton, and recognition
of the genera Metelectrona and Parvilux as valid genera on the
basis of larval characters, which Paxton had synonymized with
Electrona and Lampanyctus respectively on the basis of adult
features. Neither study restricted characters to the derived state
and the proposed phylogenies were based on overall similarities.
The present work will attempt an analysis of derived character
states and re-examine the proposed relationships within the
family.
We have used as character states (Table 63) features of adult
osteology and photophore patterns as described by Paxton (1972),
and features of larvae as described by Moser and Ahlstrom
(1970, 1972, 1974) and Ahlstrom et al. (1976) summarized in
Moser et al. (this volume). The distribution of the
character states among the genera (we have not considered sub-
genera in this analysis) is tabularized (Table 64). The criteria
for determining apomorphic character states have been consid-
ered by many, including Marx and Rabb (1972) and Zehren
(1979:153). We have used three criteria, the numbers of which
are listed after each character in Table 63: (1) Outgroup com-
PAXTON ET AL.: MYCTOPHIDAE
241
Lampanyctini
Diaphini
Triphoturus
Parvilux
Lampanyctus
Stenobrachius
Lampadena
Taaningichthys
Bolinichthys
Ceratoscopelus
Lepidophanes
Idiolychnus
Lobianchia
Diaphus
Notoscopelus
Lampichthys
Scopelopsis
Gymnoscopelus
Hintonia
Lampanyctodes
Notolychnus
Fig. 125. Phylogenetic diagram of the Myctophidae, subfamily Lam-
panyctinae. Numbers refer to the apomorphic characters described in
Table 63. Numbers in the middle of vertical lines (e.g., 4, 6) refer to
characters for which the apomorphic state is unknown. Underlined
numbers refer to apomorphic states unique to all members of a given
lineage; bracketed numbers (e.g., 59) refer to apomorphic states that
have secondanly reversed in at least one member of the lineage; non-
bracketed, non-underlined numbers refer to character states found in
all members of a given lineage but also by convergence in at least one
other taxon in the family.
parison. All previous workers have considered the Myctophidae
and Neoscopelidae as sister groups; we have taken the character
state in the Neoscopelidae to be the plesiomorphic condition
for the Myctophidae. Paxton (1972:57) described the parallel
evolutionary trends in the neoscopelids and myctophids, with
SoliYonier similar to the Lampanyctinae and Neoscopelus sim-
ilar to the Myctophinae. We have largely limited our analysis
to those characters which display only one state in the Neosco-
pelidae. Where variation occurs within the family, the character
is discussed individually below. (2) Linear photophores. We
have considered a photophore elevated out of linear series to
be apomorphic. One line of support for this decision occurs in
the ontogeny of those myctophid species with a larval PLO
photophore, which develops on the pectoral base (where it pre-
sumably has a different function from that of the adult) and
moves dorsally during development (Ahlstrom et al., 1 976:Fig.
4). Also the photophores of Neoscopelus. the only luminous
neoscopelid genus, are largely linear. However there is some
question of the homology of Neoscopelus and myctophid pho-
tophores. O'Day (1972:71) described the ultrastructure of myc-
tophid photophores and ". . . confirm(s) Brauer's ( 1 908) original
recognition of the close resemblance of photogenic tissue in the
Neoscopelidae to that found in the Myctophidae." However
Herring and Morin (1978:318) considered photophores of Neo-
scopelus and the myctophids to be very different, on the basis
Myctophini
Gonichthyini
41,42,43
Notolychnus
Krefftichthys
Protomyctophum
Electrona
Metelectrona
Symbolophorus
Myctophum
Benthosema
s^ Diogenichthys
Hygophum
Loweina
Tarletonbeania
Gonichthys
Centrobranchus
48,49
Fig. 126. PhylogeneticdiagramoftheMyctophidae, subfamily Myc-
tophinae. Numbers are defined as in Fig. 125.
of Kuwabara's (1954) description of Neoscopelus compared to
that of Brauer (1908). As ventral photophores have evolved
independently at least one other time in the stomiiform fishes
(Fink and Weitzman 1982:71), the potential for such evolution
in deeper water fishes is high enough that one cannot consider
their mere existence a case for homology. A study of the ultra-
structure of Neoscopelus photophores would be of value. (3)
Generalized larvae. The larvae of neoscopelids are highly spe-
cialized with a robust body, a large head and jaws with prom-
inent teeth, a long gut that may be coiled and large pectoral fins.
We do not think these features were present in the ancestors of
the two families, and where they are present in the myctophids,
consider they have evolved independently. We have used only
one such feature, large pectoral fins (40, Table 63) in our anal-
ysis. We consider the generalized larva of the myctophid ances-
tor had the following characters, based on the distribution of
larval features in myctophids and other teleosts: body moder-
ately slender, gut slightly S-shaped, extending to about midbody,
head moderate in size, eyes round or nearly so, without stalks
or choroid tissue, small or moderate finfold and fins and Br,
the only larval photophores present.
We have used a total of 59 characters, far fewer than the total
described in the previous studies. For many we were unable to
determine a derived state, as they displayed two or more states
or were absent in the neoscopelids. In the osteological descrip-
tions small shape differences or classifications of a continuum
were often found in both families and were not included. A
number of the characters utilized require comment or expla-
nation: (I) Jaws are long in Solhomer and short in Neoscopelus,
and following our ground rules should not be utilized. However,
they appear to be of such fundamental importance, affecting
many correlated characters and appearing to represent a major
subfamilial difference (Paxton, 1972), that they are included
here. Paxton (1972:58) considered short jaws to be primitive,
primarily because they occurred in Protomyctophum, thought
242
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 64. Character States in the Genera of Myctophidae. The 59 characters are described in Table 1.0 = plesiomorphic state, 1 =
apomoi
-phic
state, 2 -
= dif
reren
t or
adva
need
apo
mori
)hic state
9 =
unk
now:
1 or
30th
slates.
1
2
3
4
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
n
23
24
25
26
27
Krefftichthys
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
Protomyctophum
2
0
0
9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
Electrona
2
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
1
0
Metelectrona
2
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
1
0
Benlhosema
2
0
0
9
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
9
0
0
1
1
Diogenichthys
2
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
1
Hygophum
2
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
Myctophum
2
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
9
0
0
1
9
Symbolophorus
2
0
0
0
0
0
0
0
0
0
1
0
0
0
9
0
0
0
0
0
0
0
1
0
Loweina
1
2
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
2
0
0
1
0
Tarlelonbeania
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
2
0
0
1
0
Gonichlhys
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
Centrobranchus
:
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
Nololychnus
0
9
0
9
0
0
0
9
0
0
0
1
0
0
0
0
0
1
0
0
1
0
0
1
0
Lobianchia
0
0
9
0
0
0
1
1
0
1
1
0
0
0
0
0
1
0
0
0
1
0
1
1
Diaphus
9
0
0
0
0
0
1
1
0
1
1
1
9
0
9
0
1
0
0
0
0
0
0
1
Idiolychnus
0
0
9
0
0
0
1
0
0
1
1
0
1
0
0
0
1
0
0
9
9
0
1
9
Lampanyctodes
0
0
0
0
0
0
0
1
1
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
Gym noscopelus
0
0
0
0
0
0
0
0
0
0
2
1
1
0
0
1
0
0
0
0
0
0
0
1
Scopelopsis
0
0
0
0
0
0
0
0
0
0
2
0
0
1
0
0
9
0
1
0
0
1
1
1
Lampichlhys
0
0
0
0
0
0
0
0
0
0
2
0
0
1
0
0
0
0
1
0
0
0
0
1
Notoscopeliis
0
0
0
0
0
0
0
0
0
0
2
0
0
1
1
0
0
1
0
1
1
1
1
1
Hinloma
0
0
9
0
0
9
0
0
0
0
2
1
0
9
1
0
1
0
1
9
9
1
0
9
Lampadena
0
0
0
0
9
0
0
0
2
0
0
0
1
0
0
1
0
0
0
0
0
1
Taaningichthys
1
0
0
0
0
0
0
0
2
0
0
0
0
0
0
1
0
0
0
0
0
0
Ceratoscopelus
1
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
1
Lepidophanes
0
0
0
0
0
0
0
2
0
0
0
0
0
1
0
0
0
0
0
0
1
Bolinichlhys
2
0
0
0
0
0
0
2
0
0
0
0
0
1
1
9
0
0
0
0
1
Tripholurus
0
0
1
0
0
0
0
2
0
0
0
0
0
1
0
0
0
0
0
0
0
Stenobrachius
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
1
0
0
0
Parvilux
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
Lampanyctus
0
0
0
0
0
0
0
2
0
0
0
0
0
9
0
9
0
1
0
0
0
Solivomer
0
0
9
9
0
0
0
9
0
9
9
9
1
9
9
0
0
9
0
0
9
9
0
0
9
Neoscopelus
->
0
9
0
0
0
0
9
0
9
9
9
1
9
9
0
0
9
0
0
0
0
0
0
0
Scopeleng^'s
0
0
9
0
0
0
0
9
0
9
9
9
1
9
9
0
0
9
0
0
0
0
0
0
0
to represent the most primitive myctophid based on photophore
pattern. However Myers (1958) has shown that short jaws have
arisen from the long-jawed condition a number of times in
teleost ^ olution, and discussed their adaptive advantages. We
consider short jaws to be the apomorphic condition within both
the Myctophidae and Neoscopelidae, and moderate jaws also
to be derived from long jaws. (2) Extrascapulars are single in
neoscopelids; therefore two extrascapulars in some myctophids
should be the derived condition. However Paxton (1972:58)
described how the neoscopelid extrascapular differs in position
and shape from that of myctophids. Following Williston's Rule
we consider a single extrascapular to be derived from the fusion
of two elements, independently attained in each family. In Low-
eina the single condition has arisen through the loss of the dorsal
extrascapular. (4) With no outgroup with similar photophores
for comparison, we are unable to determine whether 1-2 or 3-
9 Prcs is the apomorphic state. However the two character states
follow subfamilial limits, and one of the states must be derived
and definitive for its subfamily. (6) All myctophids have at least
one of the orbital light organs, Dn and Vn, and most have both.
We are not sure whether the presence or the absence of a Dn is
apomorphic, but one of those stales defines a major line within
the Lampanyctinae. (9) Although the Neoscopelidae have a su-
pramaxillary, Paxton (1972:62) considered the supramaxillary
of some Myctophidae to be an independently derived feature.
due to a difference in shape and its required loss at least four
times within the family if considered primitive. However, John-
son (1974b:205, 1982:79) has shown the presence of supra-
maxilla(e) to be primitive in other myctophiforms (sensu lato);
the absence of a supramaxilla in myctophids is here considered
a derived state through loss. (15) Although caudal luminous
organs are not present in neoscopelids, they are present in all
but three myctophid genera, where their loss is here considered
derived. No other characters indicate that any of the three genera
(Diaphus, Gymnoscopelus, Hintoma) are the most primitive in
the family. (17) Two or three horizontal Pols are in a linear
position and should be considered the plesiomorphic condition.
However in those genera with horizontal Pols (Notoscopelus,
Lampichlhys and Scopelopsis) the photophores are high, close
to the lateral line. We consider the primitive myctophid state
to be one with low photophores with none or one Pol (character
34). We therefore consider the horizontal position of Pols to
be derived, while noting the state in Hintonia is intermediate
between angled and horizontal. (19) Although Johnson (1982:
76) considered a higher number of vertebrae (42-62) plesio-
morphic for iniomous fishes, lower numbers of vertebrae in
neoscopelids and almost all myctophids indicate the higher
number in Gymnoscopelus is a secondary specialization in these
families. (23) The larval gut of some neoscopelids is long and
coiled, clearly a specialization foreshadowing the condition of
PAXTON ET AL.: MYCTOPHIDAE
Table 64. Extended.
243
28
29
30
31
32
33
34
35
36
37 38 39 40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
1
0
0
0
0
0
0
0
0
0 0
D 0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
I
0
0
0
0
1
0
0
0
0
0
0
0
0
0 0
1 0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
9
0
1
0
0
0
0
0
0
0
9
0 1
1 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
0
9
0
1
0
0
0
0
0
9
0
1
0 1
1 0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
1
1
1
0
1
0
0
0
0
0
0
9
0 1
1 0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
9
0
0
1
0
1
0
0
0
0
0
0
0
0 1
1 0
0
0
0
2
0
0
0
0
I
0
0
1
0
0
0
0
0
1
0
1
0
0
0
0
0
2
0
9
0 1
0
0
0
0
2
0
0
0
0
0
0
1
0
0
0
9
9
9
1
0
1
0
0
0
0
0
0
0
0 1
0
0
0
9
2
9
0
0
0
9
0
0
0
0
0
1
0
0
9
0
1
0
0
0
0
0
0
1
0 1
0
0
0
2
0
0
0
0
0
0
0
0
1
0
1
0
0
9
0
1
0
0
0
0
0
0
0
0 1
1
1
1
1
0
0
1
0
1
0
0
0
0
0
0
1
0
1
0
1
0
0
0
0
0
0
0
0 1
1
1
1
1
0
0
1
0
1
1
0
0
0
0
1
1
0
0
0
1
0
0
0
0
0
0
0
0 1
1
1
1
1
1
1
0
0
1
0
0
0
0
0
1
1
0
0
0
1
0
0
0
0
0
0
0
0 1
1
1
1
1
1
1
0
1
2
0
0
0
0
0
1
1
0
1
0
0
0
0
0
0
0
2
0
0
0 0
0
0
0
0
2
0
0
0
0
2
0
0
0
0
0
0
0
0
0
9
1
0
0
0
0
0
9
0
0
0 1
1
0
0
0
2
0
0
0
0
9
0
0
0
0
1
9
0
0
1
1
9
9
0
0
0
0
9
0
0
0 1
0
0
0
0
2
0
0
0
0
9
0
0
0
0
1
0
0
0
1
9
1
0
0
0
0
0
2
0
0
9 1
9
0
0
9
2
0
0
9
0
9
0
9
0
9
9
9
9
9
1
9
0
1
0
0
0
0
1
0
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
1
0
0
0
0
2
9
0
0 1
0
0
0
0
2
0
0
0
0
9
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
2
1
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
2
1
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
2
1
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
2
0
0
9 1
9
0
0
9
2
0
0
9
0
0
0
9
0
9
9
9
9
9
0
0
0
9
1
0
0
0
I
0
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
0
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
1
1
0
2
0
0
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
1
0
0
2
1
1
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
1
2
0
0
1 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
2
0
1
0 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
9
0
0 1
0
0
0
0
2
0
0
0
0
9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
0
1 1 1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
0
0
0
2
1
1
9 1 1
9
0
0
0
2
0
0
0
0
9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
9
9 9 5
9
0
0
9
9
0
0
9
0
0
0
9
0
9
9
9
9
9
0
0
0
0
0
0
0
0
9
0
9
0 9 9
1
0
0
0
9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
9
0 9 9
1
0
0
0
9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
some acanthopterygians. Although it could be argued that the
short gut that lengthens during development in a few forms of
myctophids represents the primitive condition, we consider the
primitive myctophid condition a moderate— iengthed gut, with
different derived states, short and long. (26) Although the caudal
luminous organs are sexually dimorphic in about half the genera,
we assume the original caudal organs were not sexually dimor-
phic. (27) No photophores are present on the described larvae
of Neoscopelus. However the Br, develops in all larval mycto-
phids except Taamngichthys and Notolychmis. and its univer-
sality indicates it was present in the ancestral myctophid. Other
larval photophores however are present in fewer than half of
the genera and we consider their presence derived. (30) The
strongly developed caudal luminous organs found in Lampa-
dena and Taaningichthys are clearly a more specialized state
than the relatively unstructured organs found in many other
genera. (34) See the discussion of character 17. (36) Although a
strongly angled set of SAOs represents a linear position for the
first two photophores, we consider this condition developed by
the SAO, rising from a lower position in the weakly angled,
plesiomorphic position. (59) We consider the absence of a keel
or ridge on the fifth circumorbital of Hintonia to be secondarily
derived through loss. This is the only character state we have
used which is not present in all examined members of the line
it defines.
We have thus attempted to determine polarity for 25 osteo-
logical, 17 larval and 17 photophore characters. We initially
attempted a phylogenetic analysis utilizing the distribution of
23 larval characters at the species level. The resulting diagram
split some genera into as many as three unrelated lines. We
remain convinced that the myctophid genera as currently de-
fined by larval morph, photophore pattern and osteology rep-
resent monophyletic lines (even though such genera as Diaphus,
Lampanyctus, Myctophum and Hygophum may be formally di-
vided as subgenera or genera by future work). These genera we
use as the starting point in the present study. We have con-
structed a phylogenetic tree (Figs. 1 25, 1 26) based on our knowl-
edge of the family and used the apomorphic states of the 59
characters to define the various branching points, which is the
basis of the following discussion.
The subfamily Lampanyctinae is defined by two apomorphies
restricted to all members of the subfamily (those characters
found in all members of a lineage and nowhere else in the family
are underlined in Figs. 125 and 126). the presence of a cleithral
shelf and a single, fused extrascapular. The subfamily Mycto-
phinae is defined by two apomorphies, short or moderate jaws
and narrow larval eyes, but these features are also found in a
few genera of the Lampanyctinae. The number of Pre photo-
phores defines all members of one of the subfamilies (see dis-
cussion of character 4 above).
244
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Notolychmis valdiviae. here considered a monotypic tribe,
could not be placed with certainty in either subfamily. Moser
and Ahlslrom ( 1 970: 1 38, 1 974:409) and Paxton ( 1 972:6 1 ) dis-
cussed the characters and problems of this enigmatic species.
With long jaws and the lack of a cleithral shelf both considered
plesiomorphies, the apomorphic number of Pre photophores
unknown, and the larval eyes variable and intermediate in shape,
future work is required to resolve this trichotomy.
We recognize three tribes in the subfamily Lampanyctmae
(Fig. 125). The tribe Lampanyctini, with nine genera, is defined
by the presence of a row of moderately to strongly hooked teeth
in the posterior dentary; the only other genus with this feature
is the myctophine Diogenichthys. These nine genera are also the
only lampanyctines to lack a Dn orbital photophore, but we are
unsure if this is a derived state (see discussion of character 6
above). Moser and Ahlstrom (1972) and Ahlstrom et al. (1976:
148) placed five of these genera (Lampadena. Taaningichthys.
Bolinichthys. Lepidophanes, Ceratoscopelus) in the tribe Gym-
noscopelini, based primarily on larval photophore pattern. Pho-
tophores which appear in larvae of Lampanyctinae are essen-
tially the same ones which develop in myctophine larvae (Moser
et al., this volume) and, if they are adaptive as Moser (1981)
has suggested, it is likely that they have appeared in these typical
sites independently in a number of lineages. Moreover, these
photophores develop at the end of the larval period, if at all, in
Bolinichthys and no photophores develop in Taaningichthys
larvae. Likewise, the larval pigment characters do not support
the inclusion of these five genera in the Gymnoscopelini.
In addition to the distribution of hooked dentary teeth and
Dn photophores, other features influenced our decision about
these five genera. The ischial ligament is medium or long in all
Lampanyctini except Taaningichthys (and some species of Dia-
phus). while the fifth circumorbital has a ridge or keel in all
gymnoscopelines (but is lacking in some species of Diaphus) and
no lampanyctines except Bolinichthys (thus the brackets around
character 59 in Fig. 125). Finally all of the gymnoscopeline
genera except Notoscopelus are restricted to the southern ocean
(Moser et al., this volume: Table 59), while the Lampanyctini
are found both north and south (except Stenobrachim) of the
equator. Placement of the five genera in the Lampanyctini re-
quires fewer character reversals and parallelisms.
Within the Lampanyctini, the development of larval photo-
phores in addition to Br, (character 27) unites the five genera
discussed above. We recognize Dorsadena as a subgenus of
Lampadena until specimens other than the types are available
for osteological study and the larvae are discovered. We have
not found an apomorphic character that defines the line in-
cluding Stenobrachius. Triphoturus. Lampanyctus and Parvilux.
We are recognizing Parvilux on the basis of a weakly angled
SAO and larval shape and pigmentation.
We consider the tribe Diaphini to be the sister group of the
Gymnoscopelini. The relationships among the three genera of
Diaphini are not clear. One of us (HGM) has re-examined the
specimens on which the larval features of Idiolychmis urolampus
were based (see Moser and Ahlstrom, 1974:405-406; Nafpak-
titis and Paxton, 1978), and now thinks they could represent
Lobianchia gemellari. with the larvae of Idiolychnus still un-
known. Two characters shared by Lobianchia and Idiolychnus.
the presence of caudal organs and the absence of a luminous
patch above the pectoral fin, are considered plesiomorphic, while
the absence of a Vn and differences of photophore positions are
not clearly apomorphic. The most unequivocal derived state is
the presence of a wide pubic plate, indicating Lobianchia and
Diaphus are the sister group pair.
Within the Gymnoscopelini the proposed generic relation-
ships are based almost entirely on characters of the photophores
and luminous tissue. No consistent osteological or larval fea-
tures define generic groupings. Southern ocean larvae require
more study. The larvae of Hintonia are unknown and not enough
species of Gymnoscopelus have been studied to ascertain if the
subgenus Nasolychnus can be defined by any larval characters.
The species of Notoscopelus should also be studied to find sup-
porting characters of the subgenus Parieophus.
Within the subfamily Myctophinae (Fig. 126), we also rec-
ognize three tribes, the Electronini, Myctophini and Gonichthy-
ini. The Gonichthyini is clearly a derived lineage, with a num-
ber of osteological, photophore and larval characters
distinguishing the four genera from the rest of the subfamily.
We think the larval specializations of eyes and pectoral fins arose
after the split of the two generic pairs.
Paxton (1972) was unable to find osteological characters to
clearly separate the remaining genera of the Myctophinae into
two lineages. We have utilized photophores to distinguish the
Myctophini from the Electronini, while recognizing there is a
mosaic of osteological and larval characters within these nine
genera. We have little question of the sister group relationship
of the generic pairs Krefftichthys—Protomyctophum. Mycto-
phum — Symbolophorus and Benthosema — Diogenichthys.
However two larval features, thin head and body and a bipartite
gut, are shared by Metelectrona and some species of Hygophum.
Since we think Hygophum is a monophyletic line, we consider
these shared larval features parallelisms that do not indicate
common ancestry. Paxton (1972) considered Metelectrona a
synonym of Electrona. The description of a second species of
Metelectrona (Hulley, 1981), coupled with its larval and pho-
tophore characters, convinced us to recognize the genus.
Of the 59 derived characters utilized in our analysis, only 20
are restricted to members of the lineage they define, and eight
of these are autapomorphic at the generic level. The remaining
39 characters are not found in the apomorphic state in any
member of the opposite lineage from the defined branching
point, but are found in some members of other lineages within
the family. This presumed homoplasy of larval, photophore and
even osteological characters indicates that the proposed phy-
logeny was arrived at with some difliculty. Ten of our proposed
lineages are undefined by derived characters. We think that
future work will support our proposed phylogeny, although some
details may be modified, and that new, less plastic characters
and better definitions of polarity will help resolve the problems.
(J. R. P.) The Al'stralian Museum, 6-8 College Street,
Sydney 2000, Australia; (H.G.M.) National Marine
Fisheries Service, Southwest Fisheries Center, P.O. Box
271, La Jolla, California 92038.
Scopelarchidae: Development and Relationships
R. K. Johnson
THE Scopelarchidae has traditionally been included with the
primarily oceanic Alepisauroidei (Marshall, 1955; Gosline
et al., 1966; Rosen, 1973; Johnson, 1974b, the most recent
complete revision). Johnson (1982) excludes the scopelarchids
from the alepisauroids, rejects putative sister-group relationship
with the Evermannellidae, and provisionally allies the scope-
larchids with the chlorophthalmoids. All scopelarchids are
oceanic and meso- or bathypelagic. The majority of known adult
specimens were taken in hauls to depths between 500 and 1 ,000
m. For most species there exists no evidence to suggest diel
migration, however, Merrett et al. ( 1 973:39-40) present limited
evidence for diel migration ("considerably dispersed vertically")
in Benthalhella infans. Scopelarchids are relativedly large-bod-
ied (to 302 mm SL; Iwami and Abe, 1980). All Scopelarchidae
are tubular-eyed predators (see Munk, 1966; Locket, 1970;
Muntz, 1976; Johnson, 1982) concentrating most frequently on
fish, not capable of engorgement of enormously large food par-
ticles (unlike evermannellids, Omosudis, Alepisaurus, Antop-
terus and at least some paralepidids). Luminous tissue occurs
in Benthalhella infans (Merren et al., 1973) and probably occurs
in Scopelarchoides kreffti (Johnson, 1 974b). The family contains
1 7 species arranged in four genera and occurs throughout the
world ocean except that no scopelarchid inhabits the Arctic
Ocean or the Mediterranean Sea. Among iniomous fishes, the
Scopelarchidae is distmguished by the following combmation
of characters: ( 1 ) basihyal short to elongate but well-ossified; (2)
lingual teeth strong, straight to strongly hooked, invariably pres-
ent over basihyal, present or absent over basibranchials; (3) body
and postorbital regions of head completely covered with cycloid
scales; (4) lateral line scales large, differing distinctively in exact
conformation between all species (Johnson, 1974b: Fig. 2); (5)
parietal bones, when present, small, widely separated by frontals
and supraoccipital; (6) coracoid broadly expanded; (7) two post-
cleithra, widely separated in vertical dimension; (8) unossified
gap (filled by tube-like structure of fibrous connective tissue)
between skull and first vertebral centrum (see Merrett et al.,
1973:17); (9) posttemporal unforked; (10) no basisphenoid, or-
bitosphenoid, gill rakers, or free second ural centrum; (11) eyes
tubular, directed straight upward (except in 3 species where
directed dorsoanteriad); (12) larvae with 0, 1 or 3 peritoneal
pigment sections. The genera and species are distinguished by
gross morphological, meristic, morphometric, osteological, pig-
ment and larval characters (Tables 65 and 66).
Development
Eggs of scopelarchids are unknown. Larvae are known for all
species except Scopelarchoides kreffti and developmental series
have been illustrated and described (Rosen, 1973; Merrett et
al., 1973; Johnson, 1974b; Belyanina, 1981, 1982a;Moser, 1981).
Except for limited information on Benthalhella infans in Merrett
et al. ( 1 973), osteological description has been confined to adults.
Except in Benthalhella. development is direct, adult characters
are essentially acquired one by one, with completion of trans-
formation at 30 to more than 80 mm SL depending upon the
species. Larvae of Benthalhella undergo very rapid (i.e., small
size increment) transformation after a prolonged period of growth
while retaining larval form (see below). Larvae of most species
are known from hauls within the top 100 m and the larvae of
a number of species have been taken in the top 50 m. Con-
trariwise the larvae of one species, Benthalhella dentata. have
not been taken in hauls shallower than 150 m and most were
taken in hauls to depths in excess of 500 m. Except possibly the
cases oi Benthalhella elongata and B. macropmna (see Johnson,
1974b:228), the distributional ranges of larvae and adults are
coextensive. There is no evidence (the data are quite incomplete)
for seasonality in reproductive effort. Scopelarchids are syn-
chronous hermaphrodites.
The following paragraphs describe those characters most ev-
ident in the early life history of scopelarchids, including those
of value in distinguishing genera and species.
Gross aspect (Fig. 127). — Larvae range from extremely elongate
and shallow (Benthalhella) to quite short and deep (some species
of Scopelarchus and Scopelarchoides). Small larvae are trans-
lucent, scaleless, colorless (except for pentoneal pigment sec-
tions, when present), with a characteristic "bowed down" an-
terior dorsal profile. The body is deepest at the pectoral girdle
and the trunk elongate. Anteriorly the hypaxial muscles do not
embrace the abdominal cavity walls which are therefore highly
translucent. Only the muscles of the pelvic girdle are visibly
evident. The abdominal cavity is triangular, deep anteriorly.
Peritoneal pigment appears early except in Benthalhella which
lacks peritoneal pigment until transformation. The gut is mid-
ventral. In larvae the anus is anterior (relative to distance be-
tween pelvic fin insertion and anal fin origin) to position in
adults, far anterior in some (Benthalhella). The head is very
Table 65. Com
PARISON
OF Selected Meristic
Characters among
Scopelarchid
Species.
Lateral
Dorsal
Anal
Pectoral
line scales
Vertebrae
alalus
8-9
20-22
23-26
47-49
46-47
hubbsi
8-9
23-25
21-23
53
49
votucris
9-10
21-24
23-26
48-51
49-51
stephensi
8
20-22
18-20
41-44
42-43
michaelsarsi
7-9
18-21
18-21
40-44
40-44
anatis
7-9
21-26
18-22
45-50
44-49
guentheri
7-8
24-29
18-21
47-52
47-51
danae
6-9
24-27
20-22
50-52
48-50
nicholsi
6-7
20-23
20-23
46-50
45-48
kreffti
9
25-27
23-25
58-59
55-57
climax
7-8
25-27
25
53
49
signifer
9-10
26-29
22-25
49-52
48-49
macropmna
5-6
35-39
25-27
62-65
60-62
dentata
6-8
17-20
21-24
54-58
54-55
elongata
9-10
24-28
19-23
61-65
62-65
infans
8-9
20-26
25-28
55-59
55-58
linguidens
8-9
28-30
24-25
66
64
245
246
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
CAV
Fig. 127. Larvae, juveniles and adult of Scopelarchidae. (A. B) Rosenblattichthys volucris. A = 14.5 mm SL, B = 26.0 mm SL, letters refer to
pigment spots; (C, D) Scopelarchoides nicholsi, C = 1.5 mm SL, D = 23.0 mm SL, letters refer to larval pigment spots; (E, F) Benthalhella denlata.
E = larva, 42.8 mm SL, F = transforming specimen, 53.0 mm SL, arrows indicate position of anus; (G) Scopelarchus guenlhen. juvenile, 48.5
mm SL, DS = dermal pigment stripes; (H) Scopelarchus analis. adult, 1 12.5 mm SL.
large and massive, exceeding 30% of the SL in Rosenblattichthys,
and large but not as large in other genera. The eye is elliptically
narrowed, and initially small in comparison with the size of the
bony orbit. The interorbital is initially broad and narrows during
transformation. Development of the eyes is described for Ben-
thalhella infans in Merrett et al. (1973). The snout is pointed.
The mouth is large and low, with teeth appearing in very small
larvae. The most striking changes take place during a period of
transformation, which, as described below, can either be within
a very short interval (ca. 10 mm in Benthalhella dentata) of
growth (any statements implying time sequence are based solely
on increments of length) as in Benthalhella, or over a long (20
mm) to very long (50 mm) interval.
Meristic characters.— Counts of fin rays (Table 65) do not differ
between larval and adult specimens. Most scopelarchid species
can be uniquely distinguished from all other species on the basis
ofmeristic characters alone (Johnson, 1974b: 14). Rosenhlattich-
ihys is unique in precocious ossification of the pectoral fin rays,
well in advance of the pelvic or median fins (except caudal). In
all other scopelarchids the lowermost 5 or 6 pectoral fin rays
are the last to be formed and the order of fin ray ossification is
caudal > dorsal, anal, dorsal pectoral > pelvic > ventral pec-
toral. As in all inioms the caudal is formed of 10 + 9 principle
rays. In Scopelarchoides and Rosenhlattichthys the pelvic fins
appear as buds on the midlateral abdominal cavity wall, well
above the level of the intestine. In Benthalhella and Scopelar-
chus the pelvic fin buds appear ventrolaterally, at or beneath
the level of the intestine. In Benthalhella (except B. macropinna)
the pelvic fin insertion in larvae is distinctly in advance of the
dorsal fin origin. In other scopelarchid larvae the pelvic fin
insertion is beneath or behind the dorsal fin base (but comes to
be slightly in advance of dorsal fin origin in adult Rosenhlattich-
thys and distinctly in advance of dorsal fin origin in all adult
Benthalhella). The adipose fin develops within the dorsal finfold
which extends between the dorsal and caudal fin in small larvae.
In adults the adipose fin is inserted over the posterior one-third
of the anal fin base (except B. dentata where inserted posterior
to a vertical through base of last anal-fin ray). Ventral finfold
extending from vent to anal-fin origin in smaller larvae, and is
completely reabsorbed in early transformation.
Peritoneal pigment sections. — \n all adult scopelarchids (except
B. elongata) the gut is enclosed by a uniform tube of brown to
jet-black pigment. In larvae this pigment appears in discrete
sections (except in Benthalhella where peritoneal pigment is
lacking prior to transformation) and in a conformation char-
acteristic for each genus or group of apparently related species.
All larvae larger than 20 to 22 mm possess peritoneal pigment
(except in Benthalhella). One section only, unpaired, forming a
saddle-like canopy over the gut, is present in Rosenhlattichthys,
Scopelarchoides signifer, and S. clima.x (larvae of S. kreffii are
unknown). Three sections, a single anterior section as above and
two paired posterior sections are found in Scopelarchoides nich-
JOHNSON: SCOPELARCHIDAE
247
olsi. S. danae. and Scopelarchus. However in S. nicholsi and 5.
danae the posterior sections appear significantly "later" and
appear above (S. danae) or anterior (S. nicholsi) to the pelvic
fin bases. In Scopelarchus a.\\ 3 sections appear in near synchrony
and the posterior sections appear well to the rear of the pelvic
fin bases. In all cases the pigment section(s) expand during trans-
formation and for all genera except Benihalhella the completion
of transformation can be defined as acquisition of the adult state
of a complete and unbroken tube of peritoneal pigmentation.
In Benthalbella the first appearance of peritoneal pigment (not
in discrete section but uniformly in mesentary dorsal to gut from
between pectoral fin bases to behind pelvic fin bases) signals the
onset of the period of "rapid" transformation.
Other larval pigment.— Jht larvae of Scopelarchoides and Ro-
senblattichthys are characterized by the presence of well-defined
pigment spots or areas (accessory pigment of Johnson. 1974b;
complementary pigment of Belyanina, 1982a) apparent in the
smallest (6- 1 2 mm SL) known larvae. The presence and location
of spots is uniquely diagnostic for each species possessing them.
Pigment spots are present in all larvae of Scopelarchoides and
Rosenblattichthys. absent in Benthalbella and Scopelarchus. In
Scopelarchoides the middorsal spot, if present, and the mid-
ventral spot are entirely behind the adipose base and anal fin
base respectively. In Rosenblattichthys the middorsal and mid-
ventral (where present) spots are entirely in advance of the bases
of these fins.
Transformation pigmentation.— Johnson (1974b:20) distin-
guishes "dermal" vs "epidermal" pigmentation in scopelar-
chids. Dermal pigmentation refers to the major pigment stripes
present in some genera and species. These develop "early" dur-
ing transformation and persist in the adult. In most cases the
dermal pigment comes to be partially or completely overlain
by the epidermal pigmentation associated primarily with the
scale pockets. Dermal pigment is present in all 4 species of
Scopelarchus and in certain Scopelarchoides and Rosenblatt-
ichthys, it is absent in Benthalbella. The subequal pigment stripes
oC Scopelarchus (Fig. 127), situated above and below the lateral
line, are diagnostic for the genus.
Gut morphology.— \n all scopelarchids the stomach is a heavily
muscularized, greatly elongate blind pouch. In small larvae the
stomach does not reach the pelvic fin base, but it expands pos-
teriad during transformation, very "rapidly" so in Benthalbella.
and in all adults extends to or nearly to a vertical through the
anus (which in all is closely-adjacent to the anal fin origin).
Johnson (1974b) and Wassersug and Johnson (1976) note that
the tremendous expansion of the stomach allows ingestion of
fairly large particles and hypothesize that the blind pouch ar-
rangement is a device for maximal recovery of food energy.
Transformation. — Larvae of Benthalbella undergo rapid trans-
formation after a prolonged period of growth while retaining
larval form. The onset of transformation (size of smallest known
transforming specimen = 49.6 mm SL in B. dentata; 89. 1 mm
SL in B. elongata: 55. 1 mm SL in B. infans; 65. 1 mm SL in B.
macropinna; no transforming specimens of B. lingutdens are
known, but the largest known larva is 85.5 mm SL) is signalized
by appearance of a lens pad, appearance of peritoneal pigment,
and invasion of the abdominal body wall by musculature. Other
changes occurring during transformation include rapid elonga-
tion of gut and stomach, "migration" of anus from just behind
pelvic fin base to just anterior to anal fin origin, appearance of
gonad, appearance of scales (especially lateral line scales), ap-
pearance of head and body pigmentation, reabsorption of ven-
tral adipose fin. great restriction of base of dorsal adipose fin,
ossification of vertebral column, change (from dorsally convex
to dorsally concave) in curvature in vertical plane of anterior
portion of vertebral column (Merrett et al., 1973; Johnson,
1974b). The result is a miniature adult at the end of a trans-
formation period covering as little as 1 0 mm of growth (Johnson,
1974b:68). In other scopelarchid genera these and other adult
characters are acquired essentially one by one over an increment
of growth ranging from 15 to 50 or more mm SL [in most
transformation occurs over an actual size (SL) range of 1 5 mm
to 40 or 50 mm]. Implications of changes in morphology during
transformation in terms of activity, buoyancy, feeding and other
aspects of biology are discussed for B. infans in Merrett et al.
(1973).
Relationships
The scopelarchids were poorly known until the completion
of Johnson's ( 1 974b) revision. Currently recognized are 1 7 species
grouped in 4 genera. Phylogenetic analysis involving hypothe-
sized derived states of 1 9 characters or character complexes
(Table 66) supports allocation of species among 3 of the 4 genera.
As will be shown, Scopelarchoides remains a problem. In the
listing that follows characters are given a character number (de-
rived state number). Documentation of character state catego-
rization and hypothesized polarity are given in references listed
in the key to Table 66. Of the 19 characters for which polarity
is indicated, 6 involve larval features (Table 65: 18, 19, 20, 22,
23, 24). Of 13 adult characters, 5 represented noval autapo-
morphies (Table 65: 1,4, 13, 14, 15), 3 occur in a sequence of
3 or more steps (Table 65: 5, 11, 16), and 5 represent reductive
characters (Table 65: 6, 7, 8, 9, 12). Rosenblattichthys is dis-
tinctive in having a greatly enlarged head in larvae 19 (19) and
precocious development of the pectoral fins 20 (20). A single
reductive character 8 (7) putati vely links the remaining 1 4 species
of scopelarchids. Scopelarchus is specialized in having subequal
dermal pigment stripes above and below the lateral line 4 (2),
unique support of the first epibranchial 16(17); unique confor-
mation of the three peritoneal pigment sections 22 (24), and in
three reductive characters 9 (8), 12(11), and 23 (25). Scopelar-
chus analis is linked with 5. michaelsarsi and 5. Stephens! by
one reductive character 11 (10). Scopelarchus stephensi and S.
michaelsarsi are linked by a reduced number of vertebrae 5 (3)
and by early onset and completion of metamorphosis 24 (26).
Benthalbella is specialized in having delayed but then extremely
"rapid" metamorphosis 24 (27) and in three reductive char-
acters 6 (5), 22 (21), and 23 (25). Linking Benthalbella dentata.
B. infans, B. lingutdens and B. elongata is the unique presence
of a hooklike process on the urohyal 15 (14) and two reductive
characters 9 (8) and 1 1 (9).
In dealing with the 5 species included by Johnson (1974b) in
the genus Scopelarchoides the evidence available (Table 66, Fig.
1 28) suggests that this group is both unnatural and paraphyetic.
Linking 5. nicholsi, S. danae and Scopelarchus are unique se-
quential and fully correlated novel autapomorphies: support of
the first epibranchial character 16 (states 15 - 16 - 17), and
number and position of peritoneal pigment sections, character
22 (states 22 - 23 ^ 24). Further linking 5. nicholsi with S.
danae and Scopelarchus are relative size of the opercle and
248
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 66. Characteristics of the Scopelarchidae. Characters and character states are defined and listed below. Positive integers indicate
derived states, zeroes indicate primitive states, letters denote states of characters where polarity could not be determined.
Rosenhlalltcblh
■<
Scopelo
rchus
■s
copeiarchotdes
Benlhalhella
Char-
volu-
michael-
guenih-
nich-
sigii:-
mac ra-
Imgui-
acters
alatus
hubbsi
cris
stephensi
sarsi
a nail s
en
danae
flist
kreffn
■Imm.x
fer
pt una
denlala
ftongata
inlans
dens
Gross morphology
1
0
0
0
0
0
0
0
0
0
?1
0
0
0
0
0
1
0
2
a
a
a
b
b
b
b
b
b
a
b
b
a
a
a
a
a
3
c
a
a
b
b
b
b
c
c
c
c
b
c
c
c
c
?c
4
0
0
0
2
2
2
2
0
0
0
0
0
0
0
0
0
0
Meristic characters
5
0
0
0
3
3
0
0
0
0
4
0
0
4
4
4
4
4
Osteological characters
6
?0
?0
0
?0
0
0
0
0
0
?0
?0
0
5
5
5
5
?5
7
?0
?0
0
?6
6
6
6
6
6
?0
■'0
0
0
0
0
0
?0
8
?0
0
0
7
7
7
7
7
7
7
?7
7
7
7
7
7
7
9
0
0
0
8
8
8
8
0
0
0
0
0
0
8
8
8
8
10
b
b
b
a
a
a
a
a
a
b
b
a
b
b
b
b
b
11
0
0
0
10
10
10
9
9
9
0
0
0
0
9
9
9
9
12
0
?0
0
?11
11
11
11
0
0
?0
?0
0
0
0
0
0
0
13
0
0
0
12
12
12
12
12
12
0
0
0
0
0
0
0
0
14
?0
0
0
0
0
0
0
0
0
13
13
13
13
13
13
13
13
15
?0
?0
0
0
0
0
0
0
0
?0
?0
0
0
14
14
14
14
16
?0
?0
0
?17
17
17
17
16
15
?0
?0
0
0
0
0
0
0
Developmental characters
17
b
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
18
0
0
0
18
18
18
18
18
0
?0
0
0
0
0
0
0
0
19
19
19
19
0
0
0
0
0
0
?0
0
0
0
0
0
0
0
20
20
20
20
0
0
0
0
0
0
?0
0
0
0
0
0
0
0
21
a
a
a
b
b
b
b
a
a
?a
a
a
b
b
b
b
b
22
0
0
0
24
24
24
24
23
22
?0
0
0
21
21
21
21
21
23
0
0
0
25
25
25
25
0
0
?0
0
0
25
25
25
25
25
24
0
0
0
26
26
0
0
0
0
?0
0
0
27
27
27
27
27
KEY:
Character state classification and hypothesized polanty based on detailed information presented in Johnson (1974b. 1982) and Iwami and Abe (1980). Both characters (boldface, in brackets)
and character states (in parentheses) are numbered sequentially.
Gross morphohgy —{\\ Luminous tissue is (0) absent. (1) present; |2| pelvic-fin insertion is (a) antenor to dorsal-fin ongin. (b) prasienor to dorsal-fin origin; |31 length of pectoral fin is (a)
subequal to, (b) distinctly longer than, (c) distinctly shorter than length of pelvic fin; |4| dermal pigment stnpes as equal or subequal stripes above and below lateral line are (0) absent. (2) present.
Merislic characiers —\5\ Modal number of vertebrae. Occurs within span of (3) 40 to 44. (0) 45 to 51, (4) 54 lo 65, hypothesized character slate sequence: 3-0-4.
Oiieologica/ characlers —\(t\ An anterovcntrally directed prong from opisthotic reaching or nearly reaching border of proolic is (0) present. (5) absent; |7| panental bones are (0) present, (6)
absent; |8] supraorbital bones are (0) present. (7) absent; |9| antorbital bones are (0) present. (8) absent; |10| Ethmoid process on first infraorbital bone is (a) present, (b) absent. |ll| Supramaxillary
bones are (0) large, one-third to one-fourth the maxillary length; (9) splinllike. less than one-ninth of maxillary length, (10) absent: hypothesized character state sequence 0-9-10, [12] Discrete
postenor arm of hyomandibular bone which articulates with opercle is (0) present, (11) absent, represented only by a rounded ridge. 113) Opercle— (0) subequal to or less than, (12) distinctly
great than— subopercle in size. 1I4| (0) basibranchial teeth present, basihyal short, (13) basibranchial teeth absent, basihyal long. |I5| Hook-like process on anterodorsal surface of urohyal is (0)
absent. (14) present, |16| (0) suspensory phar>ngobranchial (PBl) present, uncinate process (UP) of first epibranchial (EBl) and second pharyngobrancial (PB2) connected by a ligament; (15)
FBI lacking, support of EBl near proximal end of PB2 — UP of EBl and PB2 connected by a ligament; (16) PBl lacking, support of EBI near middle of PB2. no UP. (17) PBl lacking, support
of EBl at point of articulation between PB2 and EB2. no UP. Hypothesized character state sequence: 0-15-16-17
Developmental characiers — |I7| Dermal pigmentation as defined in text is (a) present, (b) absent, [181 Adipose fin (0) remains elongate (extending antenad to over antenor anal-fin base)
throughout transformation penod, (18) is reabsorbed early in transformation, exhibiting adult proportions in specimens 20 to 22 mm SL and larger, II9| Head length in larvae (=28 mm SL) (0)
not exceeding 30% SL. (19) exceeding 30% SL. |20| Pectoral fin (0) not precocious, all other fins with completely differentiated rays pnor to ossification of ventralmost rays {at least) of pectoral
fin, (20) precocious, all rays completely differentiated pnor to formation of complete complement of rays of all other fins (except caudal fin), (211 Pelvic fin buds (a) form midlaterally, well above
level of intestine, (b) form ventrolaterally, at or below level of intestine, |221 Number of pentoncal pigment sections in larvae (2 1 ) = 0. (0) = I , (22) = 3, the postenor paired sections appeanng
much later in development than the single antenor section, and appeanng entirely antenor to the pelvic-fin bases. (23) = 3. the postenor paired sections appeanng much later in development
than the single antenor section, and appeanng over the pelvic-fin bases. (24) = 3. the postenor paired sections app)eanng in near synchrony with the single antenor section and appeanng entirely
posterior to the pelvic-fin bases. Hypothesized character state sequence: 21 - 0 - 22 - 23 - 24, |23| Other pigment spots or areas (as defined in text) are (0) present. (25) absent, |24|
Transformation is (26) gradual, onset at 12-14 mm SL or smaller, completion at 30-35 mm SL or smaller. (0) gradual, onset at 16-22 mm SL or larger, completion at 40-60 mm SL (most
species, R alaius is extreme with onset at 9-10 mm SL and not yet complete in 6 (39,9-80. 1 mm SL) juveniles examined by Johnson { 1 974b)], (27) abrupt; onset at 49 6-89 I mm SL or larger,
completion at 68.3-98.6 mm SL or larger (size for both onset and completion of metamorphosis vanes among the 5 species of Benlhalhella). Hypothesized character states sequence: 26 - 0 -
27.
subopercle 13 (12) and two reductive characters 7 (6) and 1 1
(9). Further linking S. danae 'wiih Scopelarchus is a unique early
restriction of the base of the dorsal adipose fin 18 (18). I am
convinced that the characters previously detailed warrant ge-
neric level recognition for the group of 4 species assigned to
Scopelarchus. Thus Scopelarchoides (type-species S. nicholsi)
should be restricted to S. nicholsi and S. danae.
This leaves the three species currently assigned to Scopelar-
choides, viz. S. signifer, S. climax, and 5". kreffti. These three
share no known derived character unique to just this group.
They share a single presumably derivative character— loss of
basibranchial teeth, extension of length of basihyal tooth row
14 (13)— with Benlhalhella but as noted by Johnson (1974b:
204) this may represent adult retention of a larval character
state common to all scopelarchids. Scopelarchoides kreffti. a
subtropical convergence species, shares with Benlhalhella an
increase in the number of vertebrae 5 (4) and probably shares
with B. infans the presence of luminous tissue 1(1). Most os-
teological characters arc unknown for 5. climax and S. kreffti
(as a resuU of paucity of available material) and the larvae of
JOHNSON: SCOPELARCHIDAE
249
-26
- 3
—25
— 24
— 17
— 11
8
— 2
— 14
9
- 8
23
—18
—16
-22
-15
-12
- 9
- 6
-20
-19
-27
-25
-21
- 5
Fig. 128. Proposed relationships among scopelarchid species based on adult and larval characters. Integers indicate derived character states,
listed in Table 66, possessed by taxa above indicated point in dendrogram.
S. kreffti are unknown. I would argue that the specializations
oi Benthalhella, especially in larval characters relating to a unique,
rapid pattern of transformation preclude addition of 5. signifer,
S. climax, and presumably 5. kreffti to Benthalhella. But with
S. climax and S. kreffti very poorly known and with the only
"character" uniting this "group" of three being that they are
"left over," I remain with my 1974b (p. 217) compromise.
Uniting all 5 species of "Scopelarchoides" and diagnostically
separating them from Scopelarchus and Benthalhella are de-
velopment and conformation of accessory pigment spots char-
acter 23, and lateral appearance of the pelvic fin bud. character
21. It is possible that the state exhibited by Scopelarchoides
larvae is primitive in both cases (I doubt that lateral appearance
of the pelvic fin buds is primitive) but until this can be shown
through adequate outgroup comparison and until S. climax and
S. kreffti are better known, I refram from attempting the de-
scription of an additional genus. Thus, for now, the possibly
paraphyletic genus "Scopelarchoides" is retained.
A summary of the contribution of 6 ontogenetic characters
to this analysis is presented below.
Dermal pigmentation (character #/7j. — Dermal pigmentation
and/or dermal pigment stripes are found in all scopelarchid
genera except Benthalhella. however, the fixation of such pig-
ment into subequal stripes above and below the lateral line is
diagnostic of and unique to the four species of Scopelarchus.
This fixation is regarded as autapomorphous for this genus.
.Adipose fin (character #18).— Scopelarchoides danae shares with
Scopelarchus an early reabsorption of most of the adipose (fin,
resulting in restriction to essentially adult proportions of the
base of this fin in specimens 20-22 mm SL. In other Scopelar-
choides as in Benthalhella and Rosenhlattichthys the adipose fin
remains elongate, to over the anterior anal fin rays, throughout
transformation, assuming adult proportions in specimens >30
mm SL. In combination with other characters uniting .S. danae
with Scopelarchus (Fig. 128) fixation of early restriction of the
dorsal adipose base is regarded as apomorphous for this group.
Head length (character #19}. — The head in larval Rosenhlatt-
ichthys is unusually large, deep and massive, the head length
exceeding 30% of the SL. The head length in other scopelarchid
larvae does not exceed 30% of the SL and this is apparently the
caseinchlorophthalmoids(Taning, 1918;Okiyama, 1972, 1974b,
1981) and most alepisauroids (Rofen, 1966a: Johnson, 1982).
Larvae of Omosudis and .Alepisaiirus do exhibit exceptionally
large heads (Rofen, 1966b). The fixation of this character in
Rosenhlattichthys alone among scopelarchids is presumed to be
apomorphous.
Pectoral fin development (character #20). — The order of fin ray
differentiation varies within and between iniomous families.
Precocious pectoral fin development is unique to Rosenhlattich-
thys among scopelarchids. It is also found in ipnopids (Okiyama,
1972, 1981) and myctophids (Moser and Ahlstrom, 1970) but
250
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
not evermannellids, Omosudis. or chlorophthalmids (Tuning,
1918; Rofen, 1966b; Johnson, 1982). It is presumed that pre-
cocious pectoral fin development in Rosenblattichthys is the
derived state.
Peritoneal pigment sections (character #22). — For an overview
of the distribution of peritoneal pigment sections in inioms see
Johnson (1982) and the account of the Evermannellidae in the
present work. The single, transverse section seen in Rosenblatt-
ichthys. Scopelarchoides climax. S. signifer and presumably 5.
kreffti is here considered the primitive state. Loss of peritoneal
pigment in the larvae of Benthalbella is clearly apomorphous.
The single and paired conformation of the 3 sections in Sco-
pelarchoides nicholsi. S. danae and Scopelarchus is unique to
this lineage among inioms. The seemingly sequential progres-
sion of states 22 - 23 - 24 (Table 66: character 22) and the
correlation of these states with states 15-16-17 of character
16 strongly reinforce the concept of monophyly for this lineage.
Larval pigment spots (character #23). — Deep-lying pigment spots
or areas occur widely among iniomous fishes (TSning, 1918;
Gibbs, 1959; Anderson et al., 1966; Rofen, 1966a; Moser and
Ahlstrom, 1970; Johnson, 1982) and their presence is here pre-
sumed to be primitive. As noted above, the position and relative
size of the spots differs between and is diagnostic of Scopelar-
choides (all 5 species) vs Rosenblattichthys.
Transformation (character #24).— Larvae of Benthalbella are
unique among scopelarchids and possibly among inioms in
achieving very large size— 50 to 100 mm or more (varying by
species) while retaining a purely larval form and then exhibiting
a very "rapid" (based on size increment relative to total size)
transformation. This pattern is regarded as autapomorphous for
this genus. Larvae of two central- water species of Scopelarchus.
S. stephensi and S. michaelsarsi. exhibit a gradual transfor-
mation typical for most inioms, but, relative to other scopelar-
chids, exhibit onset and completion of transformation at sub-
stantially smaller sizes. This is regarded as an apomorphous
feature linking these two species (as does the possibly redundant
character 5, reduction in number of vertebrae).
Johnson ( 1 982:62-10 1 ) reviews some 49 characters seemingly
related to the question of sister-group relationship of the sco-
pelarchids and evermannellids. Found were derived states in
eight characters— multiple peritoneal pigment sections, lateral
attachment of dermosphenotic, restricted insertion of RAB (Ro-
sen, 1973) muscle, reduction in number of supraneurals, and
loss of the following: sclerotic bones, antorbital bones, tooth-
plate of second pharyngobranchial and basibranchial denti-
tion—characteristic of all alepisauroids (Alepisauridae, Ano-
topteridae, Evermannellidae, Omosudidae, Paralepididae) but
not the Scopelarchidae (at least primitively). Also found were
5 derived states characteristic of the Evermannellidae + Alep-
isauridae + Omosudidae but not the Scopelarchidae. viz. pos-
session of eight infraorbital bones, reduction in number of ep-
urals and loss of the following: body scales, lateral line scales,
suspensory pharyngobranchial. Admittedly many of the features
listed are "loss" characters and thus potentially worrisome, but
why should they uniformly be absent in the groups indicated
and not in the Scopelarchidae if their correlated loss is not
indicative of relationship? On the basis of the large number of
derived states shared among alepisauroids but not shared by
scopelarchids Johnson (1982) excludes the scopelarchids from
the alepisauroids and links them (tentatively) with chloroph-
thalmoids. Only a single derived state— gap in ossification
between first centrum and the skull— links the scopelarchids
with chlorophthalmoids, but this feature is found in no alepi-
sauroid. It should be reemphasized that the characters discussed
in Johnson (1982) were specifically chosen to explore the hy-
pothesis of sister-group relationship of evermannellids and sco-
pelarchids—a notion rejected. Many additional characters need
to be studied for any rigorous analysis of iniom relationships.
It is clear that the contribution of larval characters to this anal-
ysis will be great.
Field Museum of Natural History, Roosevelt Road at
Lake Shore Drive, Chicago, Illinois 60605.
Evermannellidae: Development and Relationships
R. K. Johnson
THE Evermannellidae is one of five families included by
Johnson (1982, the most recent revision) in the primarily
oceanic Alepisauroidei. Excluded from this group are the Sco-
pelarchidae, long the supposed sister group of the evermannel-
lids, but tentatively allied by Johnson with the chlorophthal-
moids. All evermannellids are oceanic and mesopelagic,
occupying (as juveniles and adults) a wide vertical range in the
upper 1,000 m, and are not known to exhibit diel vertical mi-
gration. Evermannellids are relatively large-bodied (to 184.5
mm SL) predators, capable of engorging large food particles,
and concentrating most frequently on fish although Coccorella
may more frequently prey on squid. The family contains 7 species
arranged in 3 genera. Evermannellids are distinguished among
other alepisauroids by the following combination of characters:
( 1) an externally visible tripartite division of the tail musculature
with the epaxial and hypaxial muscles separated by a midlateral
band of muscle tissue, the lateralis superficialis; (2) lack of scales;
(3) greatly reduced, edentate basihyal; (4) restriction of gill teeth
to ceratobranchial of second arch; (5) presence of tubular or
semitubular eyes in 6 of 7 species; (6) lack of external keels on
body. The genera and species are distinguished by gross mor-
phological (eye, laterosensory pores, gut morphology, luminous
tissue), meristic, morphometric, osteological, pigment and lar-
val characters (Table 67).
JOHNSON: EVERMANNELLIDAE
251
Fig. 129. Larvae and juveniles and Evermannellidae. (A) E. balbo. showing larval phase pigmentation, D 3553 II, 8-10 mm SL; (B) E. indica.
showing juvenile phase pigmentation, ORSTOM CY III-5, 28.0 mm SL; (C) O. normalops. illustrating larval phase pigmentation and multiple
peritoneal pigment sections (shown in solid black), UH 73/8/38, 10.5 mm SL; (D) C. allantica. showing juvenile phase pigmentation, RHB 2960,
6.3 mm SL; (E) C. allantica, arrow shows location of cephalic extension of pyloric caecum, ACRE I2-18A, 25.2 mm SL (peritoneal pigment
sections not shown).
Development
Eggs of evermannellids are unknown. Larvae are known for
all species and developmental series have been partly illustrated
and described (Schmidt, 1918; Rofen, I966d; Wassersug and
Johnson, 1976; Johnson, 1982). Osteological examination has
been confined to adults. Development is direct, transformation
gradual, adult characteristics are acquired essentially one by one
but for the most part such acquisition is complete in specimens
exceeding 30 mm SL.
For all species the great majority of larval specimens has been
taken in the upper 100 m but only the larvae of three species
(Evermannella balbo, E. indica, Odontostomops normalops) have
been commonly taken in hauls to 50 m or less. The distributional
ranges of larvae and adults are coextensive and there is no
evidence (the data are very incomplete) for seasonality in re-
productive effort. Evermannellids are synchronous hermaph-
rodites.
The following paragraphs describe those characters most ev-
ident in the early life history of evermannellids including those
of value in distinguishing genera and species.
Gross aspect (Fig. /29A — Larvae and smaller juveniles of all
three genera are similar in general proportions and in having a
relatively smaller eye, smaller lens, broader interorbital, and
larger snout than larger juveniles and adults. The body is deepest
just behind the pectoral fin base. The anterior dorsal profile
descends gradually and is not bowed down. The eye in larvae
of Evermannella and Coccorella but not Odontostomops is el-
liptically narrowed, broader dorsoventrally than antero-poste-
riorly. The gut cavity is essentially triangular and quite deep
anteriorly. The snout is pointed, the mouth large, and teeth
appear in very small larvae. The most striking changes in body
proportions, in all evermannellid larvae, are correlated with the
transition from individuals with a "larval phase" pigment pat-
tern to those with a "juvenile phase" pigment pattern (see pig-
mentation, below), with the result that individuals exceeding
ca. 25 mm in the latter category are essentially miniature adults.
Meristic characters.— Counts of fin rays (Table 67) do not differ
between larval and adult specimens. The caudal is the first fin
to form, it develops 10 + 9 principal rays, as in all Aulopiformes
and Myctophiformes (sensu Rosen, 1973). Next to form, in
order, are the dorsal, pelvic, anal and pectoral fins. The pelvic
fins do not greatly change position during ontogeny, they appear
ventrolaterally beneath the posterior half of the dorsal fin and
are inserted beneath the anterior half of the dorsal fin in adults.
An adipose fin connects the incipient dorsal fin with the caudal
fin in small larvae but loses this connection and shrinks in extent
with growth of the individual, inserted over posterior one-third
of anal fin base in adults. There is apparently no variation in
the above-described features among evermannellid larvae.
Peritoneal pigment sections (Fig. 129). — In all adult everman-
nellids the gut is completely enclosed by a uniform lube of dark
brown to black peritoneal pigment. In larvae, this peritoneal
pigment appears in discrete sections. In Odontostomops there
are 12 or more peritoneal pigment sections, typically 13 to 15.
In Evermannella and Coccorella there are invariably 3 sections,
one centered over and medial to the pectoral fin insertion, one
centered (or nearly so) under the dorsal fin insertion, and one
(roughly) centered between the posteriormost pelvic fin ray base
and the anal fin origin. In all cases the sections are unpaired
and are connected broadly over the dorsal surface of the stom-
ach. In small larvae the sections form canopy-like continuous
252
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 67. Characteristics of the Evermannellidae. In the list that follows only characters useful in distinguishing evermannellid taxa are
included. Those characters also included in phylogenetic analysis are numbered; presumed primitive states denoted by 0. presumed derivative
states by integers.
Coccorella
allantica
Coccorella Everntannella Evermannelta Evermanneita Evermannella Odonloscomops
atrata ahblromi halbo tndtca megatops normatops
Gross morphology
(1) Eye (each state includes a suite of presumably cor-
related features listed in Johnson, 1982, p. 68): (0)
nontubular, (I) semitubular, (2) tubular
(2) Pyloric caecum with cephalic extension: (0) absent,
(3) present
(3) Luminous tissue, associated with ventral wall of
intestine and pyloric caecum; (0) absent, (4) pres-
ent
(4) Medial snout-pad pore (Johnson, 1982, p. 8) is: (0)
present, (5) absent
Meristic characters
(5) Dorsal fin, modal number of rays; (0) 12 or 13, (6)
10 or II
(6) Number of lateral line segments:
(0) S43, (7) £34,(8) SI8
—Anal fin rays
— Vertebrae
Morphometric characters (as thousandths of SL)
—Body depth at anal-fin origin
—Horizontal diameter of eye
— Vertical diameter of eye
— Interorbital width
— Length of longest palatine tooth
Osteological characters
(7) Basisphenoid; (0) present, (9) absent
(8) Ethmoid cartilage: (0) not forming orbital septum,
(10) considerably expanded posteriorly forming an
orbital septum
(9) Supraorbitals; (0) present, (II) absent
(10) Vertically elongate fossa centered at dentary sym-
physis; (0) absent, (12) present
(11) Jaw and palatine teeth; (0) dentary teeth in two se-
ries, at least some dentary and palatine fangs
barbed, (13) dentary teeth uniserial, all fangs un-
barbed
(12) Basihyal toothplate; (0) covers dorsal and dorsolat-
eral surface of basihyal, (14) covers only posterior
2/3 of dorsum of basihyal, (15) absent
(13) Toothplate of fourth pharyngobranchial: (0) bears
teeth, ( 1 6) edentate
(14) Toothplate of fifth ceratobranchial: (0) bears teeth,
(17) edentate
Developmental characters
( 1 5) Number of peritoneal pigment sections; (0) three,
(18) twelve or more
(16) Juvenile phase pigmentation; (19) characterized by
development of three distinct rows of very large
melanophores, each row associated with one of
three main divisions of tail musculature, (0) juve-
nile phase pigmentation not as above, with many
more melanophores and no distinct trilateral pat-
tern
1
1
2
2
2
2
0
3
3
0
0
0
0
0
4
4
0
0
0
0
0
0
0
0
0
0
5
0
7
7
S
8
8
8
0
26-30
27-29
29-32
33-37
27-31
29-31
30-35
48-50
45-47
47-49
52-54
48-52
48-50
48-52
144-191
171-210
173-200
145-181
136-173
148-162
135-170
40-65
47-65
67-81
52-72
49-93
74-85
27-42
42-76
54-70
69-87
59-81
60-97
86-110
28-40
32-47
47-61
17-26
9-19
8-20
4-17
36-52
71-96
80-100
46-69
54-69
48-73
61-69
53-69
10
10
0
0
0
0
0
0
0
0
0
0
0
11
13
13
12
12
12
12
14
15
0
0
0
0
0
0
16
0
0
0
0
0
17
17
17
0
17
17
0
18
19
19
19
19
JOHNSON: EVERMANNELLIDAE
253
sheets over the dorsal and dorsolateral margins of the gut and
these sections expand ventrad as well as longitudinally with
growth. In specimens larger than 35 to 45 mm SL the peritoneal
pigment sections coalesce to form the complete gut-enclosing
pigment tube characteristic of adults.
Other pigmentation (Fig. 129).— The major pattern of body pig-
mentation in evermannellid larvae occurs in two phases, a larval
phase and a juvenile phase, with a gradual transition between
the phases. In smaller larvae (less than 12-15 mm SL) the most
prominent body pigmentation consists of a pattern of pigment
bands arranged along the myosepta. Typically these bands are
arranged in groups (symmetrically distributed in epaxial and
hypaxial myotomal bands in the tail region, nonsymmetrical
and predominantly epaxial in the trunk region), resulting in a
characteristic barred appearance. In larvae larger than 1 2 to 15
mm SL the body pigmentation characteristic of adults begins
to appear. In Odontostomops juvenile phase pigmentation is
characterized by the development of numerous highly punctate
melanophores generally distributed over the head and body. In
Evermannella the juvenile phase is typically characterized by
the development of three rows of very large melanophores, each
row associated with one of the 3 main divisions of the trunk/
tail musculature. The median row, that associated with the lat-
eralis superficialis, is limited to the tail. Both of the other rows,
epaxial and hypaxial, extend the length of the body, from the
posterior border of the head (or nearly so) to the caudal peduncle.
In Coccorella the juvenile phase pigmentation tends to be in-
termediate in state between that of Odontostomops and Ever-
mannella. the developing melanophores tend to be larger and
more prominent than in Odontostomops. but much more nu-
merous and not arranged in rows as in Evermannella. Body
pigmentation in juveniles larger than 25 to 30 mm SL is similar
to that in adults. Development of adult pigmentation in ever-
mannclid larvae is associated with gradual (all statements im-
plying time course are based solely on size increments) disap-
pearance of the larval myoseptal pigment bands. Four of the
seven evermannellid species (Coccorella atlantica. C. atrata.
Evermannella megalops. Odontostomops normalops) are highly
melanistic as adults. In Evermannella balho. E. indica. and es-
pecially E. ahlstromi the pigmentation in adults tends to be
much more mottled, with numerous, variably-sized melano-
phores (some very large) on a light brown (in alcohol) ground
color. Obscured in adults is the longitudinal tnlateral melano-
phore pattern characteristic of juveniles.
Gut morphology (Fig. 129). — \n all evermannellids the stomach
is a heavily muscularized blind sac. The stomach expands pos-
teriad with larval growth reaching its full extension (to a vertical
just behind the pelvic fin base) in specimens exceeding 20-25
mm SL. Larvae of Coccorella are distinguished by the unique
possession of a pyloric caecum that expands anteriad with growth
and enters the head in larger larvae, juveniles and adults (Fig.
129E). The caecum is visible as a short, blind, bud-like sac on
the ventro-anterior margin of the intestine in the smallest known
larvae of Coccorella. Wassersug and Johnson (1976) describe
in detail the structure and development of this remarkable or-
gan. Neither Evermannella nor Odontostomops nor (as far as is
known) any other alepisauroid possess a pyloric caecum.
rra«s/orwa?/o/i. — Development of juvenile phase pigmentation
signals the onset of transformation in all evermannellid larvae.
2,8,12,19
Fig. 1 30. Proposed relationships among evermannellid species based
on adult and larval characters. Integers indicate derived character states,
listed in Table 67, possessed by taxa above indicated point in dendro-
gram.
Transformation in Evermannellidae is gradual, adult characters
are essentially acquired one by one, and there are no abrupt and
radical changes in morphology. In all evermannellid species,
individuals larger than 25 to 30 mm SL are (except for final
fusion of pentoneal pigment) essentially miniature adults and
can be distinguished readily on the basis of adult characters
(e.g., eye morphology, presence or absence of dentary fossa,
posterior extent of lateral line, arrangement of cephalic latero-
sensory pores, dentition, pigmentation, meristic and morpho-
metric characters). Final fusion of the peritoneal pigment sec-
tions occurs by about 35 mm SL (Coccorella. Evermannella) or
by about 45 mm SL (Odontostomops).
Relationships
The evermannellids were poorly known until the completion
of Johnson's (1982) revision. Currently recognized are 7 species
in 3 genera (Fig. 130). Phylogenetic analysis involving presum-
ably derived states of 1 6 characters or character complexes sup-
ported previous allocation of species among the 3 genera. In the
listing that follows characters are given as character number
(derived state number). Of the 1 6 characters, 2 involved larval
features (Table 67: 15, 16). Of the 14 adult characters, 5 rep-
resented novel autapomorphies (Table 67: 2. 3, 8, 10, II), 3
exhibited a sequence of 3 steps (Table 67: 1,6, 12) and 6 rep-
resent reductive characters (Table 67: 4, 5, 7, 9, 13, 14). Odon-
tostomops is specialized in having 12 or more serially arranged
pentoneal pigment sections 15 (18) and in two reductive char-
acters 7 (9) and 9 (11). Coccorella exhibits autapomorphies in
four characters: cephalic extension of pyloric caecum 2(3), pres-
ence of luminous tissue 3 (4), posterior expansion of ethmoid
cartilage 8 (10), arrangement and morphology of dentary and
palatine teeth II (13) and is apomorphous in two additional
reductive characters 12 (14) and 14 (17). Coccorella atrata is
apomorphous in two reductive characters, 12(15) and 13 (16).
Linking Coccorella and Evermannella are intermediate states
in the two 3-step characters 1(1) and 6 (7). Evermannella shows
254
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
autapomorphies in three characters: unique pattern of juvenile
phase pigmentation 16 (19) and presence of vertically elongate
fossa at dentary symphysis 10 (12), presence of fully tubular
eye, 1 (2), unique to them among evermannellids, and show
further reduction in the number of lateral line segments 6 (8).
A single reductive character 14(17) also shared with Coccorella
links E. indica and E. ahlstromi and E. megalops. A final, ques-
tionable character 5 (6) links the latter two. In each case well-
defined autapomorphous features support the hypothesis of
monophyly of each genus and the information available appears
to adequately support most of the proposed scheme.
Details concerning the contribution of two larval characters
to this analysis are discussed below.
Peritoneal pigment ^frt/0/15. — Discrete peritoneal pigment sec-
tions are striking features of most aulopiform but not mycto-
phiform fishes (Johnson, 1974b, 1982; Okiyama, 1974b, this
volume). A single dorsomedial section characterizes the larvae
of all Aulopus (Okiyama, this volume), chlorophthalmoids and
(primitively) scopelarchids. Multiple (3 or more, serially ar-
ranged, paired or unpaired) sections occur in ipnopids (Bathy-
pterois), bathysaurids, synodontids, harpadontids, paralepidids,
Oinosudis and evermannellids. Peritoneal pigment sections are
paired, left and right, in synodontoids (sensu Johnson, 1982)
but single and connected dorsomedially over the gut in alepi-
sauroids. Peritoneal pigment sections are apparently lacking in
notosudids, some ipnopids, Alepisaurus, neoscopelids (perito-
neal pigment present but not in a discrete section, see Okiyama,
this volume) and myctophids. Johnson (1982) concludes that a
single dorsomedial section is primitive for aulopiform fishes.
Three unpaired sections are found in larvae of Coccorella, Ev-
ermannella, Omosiidis and the paralepidine barracudina Par-
alepis atlantica (said by Rofen, 1966a:238, to be ". . . the most
primitive species in the Paralepididae."). Larvae of Odontosto-
niops norinalops exhibit 1 2 or more unpaired peritoneal pig-
ment sections, unique in the order, and a feature regarded as
autapomorphous.
Juvenile phase pigmentation— ]o\\nson (1982) regarded fixa-
tion of the trilateral longitudinal pattern of juvenile phase pig-
mentation, as described above, as autapomorphous for the genus
Evermannella.
It has long been supposed (Gregory and Conrad. 1936; Mar-
shall, 1955; Gosline et al., 1966) that the Scopelarchidae and
Everrriannellidae are closely related. This supposition was based
mainly on the occurrence of tubular eyes in both groups. John-
son (1982) argues against this notion, rejecting any close rela-
tionship of the Evermannellidae and Scopelarchidae, placing
the latter (tentatively) among the chlorophthalmoids, and plac-
ing the Evermannellidae as the sister group of the Omosudidae
plus Alepisauridae. The evidence for these conclusions is pre-
sented in Johnson ( 1 982) and briefly summarized in the account
of the Scopelarchidae in the present work.
Field Museum of Natural History, Roosevelt Road at
Lake Shore Drive, Chicago, Illinois 60605.
Myctophiformes: Relationships
M. Okjyama
IN the traditional concept, the order Myctophiformes is con-
sidered to be a monophyletic assemblage with taxa having
much the same levels of organization, even though they have
undergone considerable adaptive radiation including some ex-
tremely specialized forms for particular habitats (Goody, 1969;
Marshall and Staiger, 1975; Johnson, 1982).
Modem definition of the order including 16 families was first
established by Gosline et al. (1966). They recognized the fol-
lowing two suborders:
Myctophoidei: Aulopidae, Synodontidae, Bathysauridae,
Harpadontidae, Bathypteroidae, Ipnopidae, Chlorophthal-
midae, Myctophidae and Neoscopelidae.
Alepisauroidei: Notosudidae (=Scopelosauridae), Paralepidi-
dae, Omosudidae, Alepisauridae, Anotopteridae, Ever-
mannellidae and Scopelarchidae.
This dichotomous system has been generally followed by re-
cent workers (Rosen and Patterson, 1969; Marshall and Staiger,
1975; Sulak, 1977), despite some minor changes or disagree-
ments in the definition of family limits. On the other hand,
Gosline (1971) proposed the idea of splitting the order into four
groups (!) without giving rigorous evidence.
Rosen (1973) reevaluated the relationships among the Myc-
tophiformes and produced a very different provisional classi-
fication based on a cladistic analysis of the group, where all of
the myctophiform fishes (except Myctophidae and Neoscopeli-
dae) form a monophyletic group, and likewise all the alepisau-
roid families (except Giganturidae) form a monophyletic assem-
blage. His phyletic hypothesis is radically different from those
of Gosline et al. (1966) and Johnson (1982).
Notosudidae was later transferred from Alepisauroidei to
Myctophoidei (Bertelsen et al., 1976), and furthermore, Sco-
pelarchidae was removed from Alepisauroidei (sensu lato) in
the recent study of Johnson (1982) who further subdivided the
order into five possible major groups in three perceived lineages
(Fig. 131).
Among these studies, Johnson (1982) is unique in carefully
evaluating larval characters such as the peritoneal pigment sec-
tions and the stomach pigmentation in juveniles, in considering
myctophiform phylogeny with special references to Scopelar-
chidae and Evermannellidae.
As finely reviewed by Kendall ( 1 982), myctophiforms provide
an excellent example for elucidating systematic relationships
among fishes using larval characters, because larvae are known
for representatives of most of the families and in some cases
nearly all of the species within the families. Potential usefulness
of the larval groups in this connection has been well documented
for several families such as Myctophidae (Moser and Ahlstrom,
OKIYAMA: MYCTOPHIFORMES
255
1
AULOPIDAE
< w w — O t^ -•
TTtH myctophidae
ttH neoscopelidae
y
ttttHNOTOSUDIDAE
I nil I SCOPELARCHIDAE
"pCHLOROPHTHALMIDAE
ttH IPNOPIDAE
— 3
TTT
TTT
tH synodontidae ^
harpadontidae
TT
T-i bathysauridae
^
TTTTTTTnr
paralepididae
N
II iiMiii II I ANOTOPTERIDAE
Mill
-tH EVERMANNELLIDAE
St
TTTTTrr
tHOMOSUDIDAE
ttH ALEPISAURIDAE
J
Fig. 131. Possible interrelationships among myctophiform fishes (Johnson, 1982).
1972, 1974), Scopelarchidae (Johnson, 1974b, 1982), Notosu-
didae (Bertelsen et a!., 1976) and Evermannellidae (Johnson,
1982). At higher taxonomic levels. Okiyama (1974b, 1979b,
1981) considered the relationships among families with partic-
ular reference to the peritoneal pigment sections in association
with the meristic features of the axial skeleton, notably precau-
dal and caudal vertebrae. Larval characters of possible system-
atic importance among Myctophoidei in Okiyama ( 1 979b) have
been closely analyzed by Kendall (1982) in establishing familial
interrelationships on the basis of the cladistic method, although
several larval stages critical to this were not available at that
time.
Since current knowledge reveals slightly different conclusions
for larval characters of potential phylogenetic importance from
those employed in Okiyama ( 1 979b), some comments are given
below for a revised character catalogue with a discussion of
possible evolutionary direction. The determination of this di-
rectional change is generally based on the assumption that the
family Aulopidae, as presently considered, represents the prim-
itive character state.
In the following discussion, the character states believed to
be primitive are all identified with a "0," and those believed to
be derived are designated by a positive integer.
Peritoneal pigment sections {1).—The development of the dis-
crete peritoneal pigment sections is a remarkable feature of lar-
val myctophiform fishes. Nothing is known of their function,
but the systematic importance of this unique structure has been
256
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 68. Distribution of Larval Character States Among
Myctophiform Families.
Characters
Family
1
T
3
4
5
6
7
8
D*
Aulopidae (Au)
0
0
0
0
0
0
0
0
0
Myctophidae (My)
3
1
0
1
0
0
0
0
3
Neoscopelidae (Ne)
3
1
0
1
0
1
2
0
5
Chlorophthalmidae (Ch)
0
0
0
0
0
0
0
0
0
Ipnopidae (Ip)
0
0
1
1
0
0
0
0
2
Notosudidae (No)
3
0
0
0
1
0
1
0
3
Scopelarchidae (Sc)
0
1
0
0
1
0
0
0
->
Bathysauridae (Ba)
1
1
t
0?
0
0
1
1
5
Harpadontidae (Ha)
2
1
0
0
0
0
1
0
3
Synodonlidae (Sy)
2
1
0
0
0
0
1
0
3
Alepisauridae (Al)
3
0
1
1
0
1
2
6
Anotoptendae (An)
3
0
0
0
0
0
1
3
Evermannellidae (Ev)
1
1
0
0
1
0
->
5
Omosudidae (Om)
1
0
0
0
0
1
2
4
Paralepididae (Pa)
1
0
0
0
0
0
1
3
* Number of denved character slates.
repeatedly emphasized (i.e., Okiyama, 1974b, 1979b, 1981.
Johnson, 1974b, 1982). Contrary to earlier understanding (Oki-
yama, 1974b), much diversity of this character has been re-
vealed. Based on the number and shape of the sections, a pro-
visional classification is as follows: (A) Dorsomedial pigment
sections; (A-1) Single patch— Aulopidae, Chlorophthalnms.
Bathytyphlops. Rosenblattichthys. Scopelarchoides (in part); (A-
2) Many (three or more)— Bathysauridae, Bathypterois (in part),
Sudis. Omosudidae, Evermannellidae; (A-3) Single to many
patches with growth— Paralepididae (except Sudis). (B) Paired
pigment sections— Harpadontidae, Synodontidae. (C) Dorso-
medial and paired pigment sections— Scopelarchoides (in part),
Scopelarchus. (D) No pigment sections— Neoscopelidae (except
Solivomer), Myctophidae, Ipiiops. Bathymicrops. Bathypterois
(in part), Benthalhella. Notosudidae, Alepisauridae, Anotop-
teridae.
Rare exceptions are also known for several of these types.
The only known exception to the presence of the A-3 type in
paralepidids is in Notolepis coatsi with a single pigment section
throughout all stages (Efremenko, 1978, 1983a). However, the
ontogenetic development of this section into the extensive per-
itoneal pigment tube around the gut as in other paralepidids
reveals little phylogenetic difference for this exception. Among
those having B-type, some Synodus reportedly lack the peri-
toneal pigment sections and may represent an extremely spe-
cialized character state (Cressey, 1981). On the contrary, a my-
tophid, Protomyctophuin anderssoni. is known to develop the
serially arranged paired pigment patches similar to those of B-
type(MoserandAhlstrom, 1974; Efremenko, 1976). Their over-
all resemblance including this pigmentation may be a result of
a simple convergence.
As is clear from the above classification, character states are
remarkably diverse in the Scopelarchidae and Ipnopidae. C-
type, peculiar to the former, is of particular significance in sug-
gesting the possible direction differentiating the paired and un-
paired character states (Johnson, 1974b). Unclear limits of the
family are partly responsible for the confusion in Ipnopidae.
It is generally agreed that the presence of a single, dorsomedial
peritoneal pigment section (A-1 type) represents the primitive
state. Since A-3 and C types are referable to the ranges of either
A-1 or 2, four states are recognized as in Johnson ( 1 982); (0) =
A single, dorsomedial peritoneal pigment section. (1) = Multi-
ple (3 or more), serially arranged, unpaired peritoneal pigment
sections. (2) = Multiple (3 or more), serially arranged, paired
peritoneal pigment sections. (3) = Peritoneal pigment section
absent.
Position of anus f2A— Contrary to the usual pattern of the anus
location immediately anterior to the anal fin origin in much
teleosts, a more or less wide preanal interspace is commonly
shared by many taxa of this order. This character can be of much
use in distinguishing the groups of Myctophiformes (Rosen,
1 97 1 ; Okiyama, 1 979b). The character states of the diverse anus
location relative to the pelvic fins are not recognized herewith
due to the unclear patterns of occurrence.
The ontogenetic rearward shift of the anus is restricted to
some speciose families such as Scopelarchidae, Paralepididae
(except Sudis), Notosudidae, and Myctophidae (in part). There
is, however, a sharp contrast in the final condition among them:
no preanal interspace in Myctophidae and Scopelarchidae vs a
distinct space in the remaining two. As in Kendall ( 1982), who
employed this character in the first step of branching, two char-
acter states are recognized. (0) = Anus with interspace from the
origin of the anal fin. (1) = Anus just in front of the ongin of
the anal fin.
Fin features (3). — Except Bathysauridae with magnificently en-
larged fins, the elongated pectoral fins are the pronounced larval
character found in many representatives of this order. The Ip-
nopidae displays the most diverse pattern of specialized pectoral
form in being bifid, large and fan-like, or extremely elongated.
Parallel features are known to occur sporadically in some spe-
cialized myctophidae (Moser and Ahlstrom, 1970, 1974). Sco-
pelarchidae is another member of less cohesiveness in this char-
acter; prominent pectoral fins are peculiar to Rosenblattichthys,
the most specialized genus in this family (Johnson, 1974b).
Likewise, only the aberrant genus Sudis has elongated pectoral
fins in Paralepididae. The character states recognized are: (0) =
All fins short. (1) = Only pectoral fins elongated. (2) = All fins
elongated.
Sequence of fin formation i^-^A- Although current knowledge is
far from complete, dichotomous patterns can be recognized in
the sequence of fin formation, especially in the pectoral fins.
Johnson (1982) defined the derived character state of Rosen-
blattichthys by the development of pectorals prior to all fins
except caudal. The precocious nature of this fin apparently rep-
resents the derived state. (0) = Pectoral fins not precocious. (1) =
Pectoral fins precocious.
Eye shape (5). — Moser and Ahlstrom (1974) showed that two
types of eyes, i.e., round and narrow, reflect the major two
lineages of Myctophidae with several exceptions. These patterns
are commonly duplicated at familial levels in this order. In view
of the specialized morphology of the narrow eyes including the
peculiar choroid tissue and following the suggested phylogeny
of Myctophidae (Moser and Ahlstrom, 1974), round or nearly
round eyes are regarded primitive. The states recognized are:
(0) = Eyes rounded or nearly rounded. (1) = Eyes narrowed.
Head armature (6).— The development of head spines is rare
or rather exceptional particularly in the adult myctophiform
fishes. However, larvae of at least five families have head ar-
mature. These include preopercular spines and supraorbital and/
OKIYAMA: MYCTOPHIFORMES
257
Table 69. Similarity Matrix of 1 5 Families of Mvctophiformes. Based on the total number of characters shared in the same state regardless
of the primitive or denved (below the diagonal) and that of the shared derived states (above the diagonal, with similarity index in parentheses).
Subordinal groups are indicated by enclosure. Similarity index is calculated by the following formula: P„ = (C,/\/S,S,) x 100, where S, and S,
are number of derived characters in families i and j, and C„ is number of the shared derived states between the same set of families.
My
Ip
Sy
Au
-
0
0
0
0
0
0
0
0
0
0
0
0
0
0
My
5
-
|3(77)|
0
1 (40) 1 (33)
1 (40) 1 (26)
1(33)
1 (33) 2
(47)
1(33)
1(26)
0
0
Ne
3
g
[6]
5
3
0
1(22
) 1 (26)
1 (32)
1(20)
1(26)
0
1 (26) 1 (55)
0 0
1(26)
0
2(40)
0
2(45)
0
0
Ch
0
0
0
0
0
Ip
6
5
3
6^^
\^"-^
0
0
1(32)
0
0 2
(58)
0
0
0
0
No
5
6
3
4
5
3
3
3
2
5
6
3\
4
^
1(45)
1(20)
1(33)
1(41)
1 (33) 1
24)
2(67)
0
1(26)
1(26)
2(64)
1(20)
0
0
0
1 (33)
So
-
1(32)
1(41)
0
0
Ba
3
2
2
2
[2^2)
2(52)
0
1 (26)
Ha
5
5
2
5
5
4
4
4
5
5
5
2
3
3
4
5
5
2
5
5
5
"^
3(100)
0
1(33)
1(33)
1(26)
1(26)
0
0
1(33)
Sy
-
0
1(33)
A!
0
2
1
1
12^7)
2(37)
3(61)
1 (24)
An
5
4
3
5
3
6
3 5
5
5
4
~~"\
,^^^^
^26)
1(29)
2(67)
Ev
3
3
3
3
2
4
5 5
5
5
2
4"^
,^^^^^.
3(67)
2(58)
Om
4
5
2
3
4
2
4
5
2
3
3
4
2 4
3
5
3
5
3_
5
7
5
"^
2(52)
Pa
3
6
5
-
or frontal ridges. Development of head armature generally oc-
curs in the forms with a massive head more than 30% of body
length, thus suggesting the specialized condition of this char-
acter. Several myctophid species {Lampanyctus) having pre-
opercular spines provide a fine example of this trend, while this
is not the case in Scopelarchidae. According to Nafpaktitis(1977),
the character state of Neoscopelus is assigned to the Neosco-
pelidae. The states recognized are: (0) = Head armature absent.
(1) = Head armature present.
Body shape f/A — The general body shape can range from ex-
tremely slender and elongate to stubby and deep. These are
tentatively grouped into three character states with possible evo-
lutionary trends towards the opposing directions from the mod-
erately slender body shape shared by primitive groups such as
Aulopidae and Chlorophthalmidae. The character states rec-
ognized are: (0) = Body moderately elongate. (1) = Body ex-
tremely slender and elongate. (2) = Body stubby and deep.
Pigment spots or area ("(SA— Johnson (1982) suggested the po-
tential importance of pigmentation other than the peritoneal
sections in the systematics of the Myctophiformes, even at high
taxonomic levels. A difficulty in this regard is how to recognize
the meaningful character states. Based on the various pigmen-
tation patterns in the tails of larvae (posterior to the anus except
for the caudal fin) such as (a) absent, (b) present along only the
ventral midline, and (c) present along lateral or dorsal surfaces
of body sometimes forming clear bands, formal recognition of
this character is undertaken. Since patterns (a) and (b) are shared
commonly during the ontogeny of the same species, two char-
acter states are recognized with the assumption that (c) repre-
sents the derived state. (0) = Pigment spots or areas in tail absent
or present along only the ventral midline. (1) = Pigment spots
or areas in tail present along lateral or dorsal surface.
The primitive or derived states for these eight characters are
summarized in Table 68. Family level designation of character
states is mostly based on the assumption of Johnson (1982) that
"possession by one or more representatives of a particular OTU
of a state considered primitive indicates (except where contrary
evidence can be cited) the primitiveness of that state for that
OTU."
A similarity matrix based on the total numbers of characters
shared in the same state, regardless of whether the states are
primitive or derived, is given below the diagonal in Table 69.
Above the diagonal are shown the numbers of derived characters
shared in the same state and the similarity index calculated on
the same data. These two sets of figures are expected to reveal
certain clues to clarify the interfamilial associations of this order
from the larval standpoint.
AiiLOPOiDEi: Aulopidae
So far as the selected larval characters are concerned, the
Aulopidae can not be separated from the Chlorophthalmidae.
This unclear distinction is due to the limited numbers of char-
acters selected, because other larval and adult features shown
in Table 70 reveal the trenchant differences between them. Of
these, the possession of maxillary teeth and fulcral scales, and
the earlier differentiation of the peritoneal pigment spots well
justify the distinct and less specialized systematic status of the
Aulopidae. Other aspects of sharp contrast such as in the den-
tition, particularly of the basihyal, and gut morphology sub-
stantiate the above conclusion.
Although the diversity within the Aulopidae once suggested
on the basis of larval characters (Okiyama, 1974b) has proved
to be unacceptable, there still remain problems concerning the
monotypic nature of this family. As mentioned elsewhere (Oki-
yama, 1979b), it is likely that the Myctophiformes evolved along
several lines, one of the major trends being the elongation of
the body shape accompanying an increase in vertebral number.
Obviously, aulopids lie near the base of this trend with clear
orientation toward an increase in the number of abdominal
components. The uniquely elongated larval oesophagus in A.
258
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 70.
Anatomical Differences of Early Stages Betwfen
AuLOPvs and Chlorophthalmus.
Table 71.
Distribution of Larval Character States among Four
Genera of the Ipnopidae.
Autopus
Chtorophlhatmits
(afler Rosen (1971)
and Sulak(1977)]
D«
Maxillary teeth
Vomerine teeth
Basihyal
Fulcral scale
Gut morphology
Peritoneal pig-
ment sections
Present
Only two widely sepa-
rated at opposing
anterolateral comer
Ovoidal with slightly
indented tip; teeth
absent
Present
Moderately elongated,
straight; intestine
slightly fat
Single; distinct at less
than 3.5 mm SL
Absent
Transverse row of six
teeth divided into
two rows of three
each
Triangular with similar
anterior indentation;
a transverse row of
six teeth divided into
two series
Present(?)
Short, compact with
slender stomach; in-
testine fat
Single; distinct at more
than 5 mm SL
japonicus is a probable indication of this evolutionary trend
(Okiyama, 1974b). Among recent congeners, A. damasi may be
the most generalized species in view of its smallest number of
vertebrae (20 + 16) similar to the known counts in the fossil
aulopids (Goody, 1969; Rosen and Patterson, 1969). Further-
more, this species is clearly separable from congeners by the
mode of direct association between the first haemal spine and
anal pterygiophores (Okiyama, 1979b). A look at the larvae of
A. damasi would be enlightening in clarifying the problem in
question.
Myctophoidei: Neoscopelidae,
Myctophidae
The two families of this suborder are readily discriminated
from the others by the greatest similarity index value based on
a suite of derived characters (1 and 2) not shared by any other
families. The smaller sizes at metamorphosis are also peculiar
to these families. These larval evidences offer strong support
for the views of Moser and Ahlstrom (1974) and Johnson (1982),
warranting a distinct subordinal ranking. My observation of the
vertebrae of Solivomer (see Table 57 in my Myctophiformes:
Development, this volume) also disclosed their closer linkage
than assumed by Johnson (1982).
The similarity matrix in Table 69 would offer little support
for Rosen's scheme to transfer these families to a different order.
Chlorophthalmoidei: Notosudidae,
Scopelarchidae, Chlorophthalmidae,
Ipnopidae
The larval character states indexed in Table 7 1 are less prom-
ising in support of this familial assemblage, because only the
Notosudidae and Scopelarchidae share a single derived char-
acter state (narrow eye). It seems that this ambiguity is also
associated with the inadequate numbers of characters in ques-
tion.
Although the admitted cohesiveness of larval characters of
Chlorophthalmidae may be altered by the discovery of larval
Bathysauropsis or Parasudis. larval characters support the tra-
ditional view that it is one of the basal stocks of this order, lying
Bathytyphlops
Ipnops
Bathymicrops
Bathypterois
10 0 0 0 2
10 0 10 4
10 0 10 4
10 0 0 15
' Number of denved character slates.
at a somewhat advanced place along a line different from the
Aulopidae. Trenchant characters in this connection such as the
dentition and the mode of anal fin support are shared with the
Ipnopidae.
Members of the Notosudidae, the most cohesive family in
this suborder, have the greatest numbers of derived characters
of the group. Marshall ( 1 966a) and Bertelsen et al. ( 1976) stated
that it seems most closely related to Chlorophthalmidae. The
superficial resemblance of larval stages between this and the
Paralepididae was also suggested (Ahlstrom, 1972a). On the
other hand, the similarity matrix indicates its affinity with An-
otopteridae, along with Scopelarchidae. Of these associations,
the last grouping based on a single derived state in character 5
(narrow eye) appears less arguable. Other features such as the
maxillary teeth and the uncommon morphology of the corpus
cerebelli suggest the aberrant systematic status of this family.
Since Table 68 provides few clues to discuss the confused
family limits of the Ipnopidae, the same coding of the character
states is applied to the four genera of this family (Table 71).
Except for the distinct larval status of Bathypterois. derived
characters shared among the remaining three genera do not
reveal the generic linkages suggested by Sulak (1977). By the
same reasoning as discussed before concerning the relationships
between Aulopidae and Chlorophthalmidae, the derived state
in character 1 (peritoneal pigment sections) shared by Ipnops
and Bathymicrops includes the different states of gut morphol-
ogy. It seems these genera form a loose but distinct assemblage
warranting family rank. Besides the shared dentition mentioned
before, the close fit of general larval morphology between Bathy-
typhlops and Chlorophthalmus may suggest their relationship.
The diverse larval characters of Scopelarchidae were elabo-
rately enalyzed in the light of adult systematics (Johnson, 1 974b).
It is remarkable that this family has no phenetic similarity with
Alepisauridae in terms of catalogued characters. On the other
hand, two derived states in character 2 (anus location) and 5
(eye shape) shared with Evermannellidae give the greatest sim-
ilarity index value. Johnson (1982) suggested the independent
occurrence of the tubular eyes in adults of both families, but
traditional concepts of their close association should be reevalu-
ated using larval evidence.
Synodontoidei: Bathysauridae,
Harpadonti[5ae, Synodontidae
Accepted linkage between Synodontidae and Harpadontidae
is clearly substantiated by the larval characters, while familial
allocation of Saurida remains to be solved. Synodus lucioceps,
having the intermediate state of larval characteristics between
these families, may be important here. The relationships among
four genera are thus indistinct from the standpoint of the larvae,
but Saurida appears to be the most generalized. Possible phy-
OKJYAMA: MYCTOPHIFORMES
259
logenetic association between Aulopidae and these families has
been suggested on the basis of larval characters and the similar
mode of anal fin support (Okiyama, 1974b, 1979b). To these
can be added the peculiar structures on the chorion surface of
the extremely transparent eggs, the pigmentation patterns in the
newly hatched larvae, and the mode of reproduction shared by
these families, characters which favor their close association.
Bathysauridae is distinguished from other families of this
suborder by some trenchant differences in the peritoneal pig-
ment sections and the mode of reproduction, while two derived
states are shared by all families. The phylogenetic relationship
of these families depends on whether the above mentioned dif-
ferences are due to divergence. Larval stages of Bathysauridae
are surely highly specialized, adapting to a prolonged pelagic
life, but larval dentition described in detail by Rosen (1971)
and Johnson (1974) and the character state of the axial skeleton,
including the mode of anal fin support (Okiyama, 1976b) are
of particular interest in showing the pattern common to Ipno-
pidae.
Alepisauroidei: Paralepididae,
Anotopteridae, Evermannellidae,
Omosudidae, Alepisauridae
The similarity matrix provides certain indication of the co-
hesiveness of this suborder. Most remarkable is their common
sharing of the derived state of character 8. Regarding the per-
itoneal pigment sections dividing five families into two groups,
some comments are warranted for Alepisauridae. As discussed
by Johnson ( 1 982), this character state is very tentatively defined
due to the inadequate state of available material. Even so, a
distinct family pair of Alepisauridae and Omosudidae can be
readily separated from the remaining families by the many de-
rived character states shared by them. Although the possibility
of their convergence cannot be fully rejected in view of the clear
contrast in the ontogenetic aspects of the pectoral fins, the close
similarity between Alepisaurus ferox and Omosudis lowei (trop-
ical western Pacific specimen) (see my Myctophiformes: De-
velopment, Fig. 1 1 2B, E, F, this volume), in head armature and
pigment pattern is extremely striking.
An association between the Anotopteridae and Paralepididae,
particularly the more elongated paralepidids such as Stemo-
nosudis and Macroparalepis (Rofen, 1 966a, c), can be seen from
the larval standpoint. In addition to their shared derived char-
acter states (character 7 and 8), a fleshy projection on the lower
jaw tip peculiar to Anotopteridae and Stemonosudis macrura.
and the similar larval dentition (huge canines) may substantiate
the above association. Their disagreement in the character of
the peritoneal pigment sections is probably associated with the
odd systematic position of Anotopteridae lying at "an extreme
specialized end-point of the paralepidid line" (Rofen, 1966a, c).
On the basis of the larval characters, two subfamilies of Par-
alepididae are well separated. As compared with the relative
constancy of conservative characters in larval Paralepidiinae,
the many derived character states of larval Sudinae are too
specialized to be consistent with the accepted subfamilial level.
The latter may be an earlier offshoot preceding the remarkable
paralepidine radiation. The complete lack of intermediate forms
between them offer strong support for this suggestion.
As in Scopelarchidae (Johnson, 1974b), the systematics of
Evermannelidae were studied in detail using a large character
suite, including larval aspects (Johnson, 1982). So far as the
present analysis is concerned, this family seems variously as-
sociated with families of Alepisauroidei such as Paralepididae,
Alepisauridae and Omosudidae, besides Scopelarchidae. It is of
interest that limited character states shared by Evermannellidae
and Alepisauridae are restricted to derived ones, probably sug-
gesting their close association. Perhaps, an Evermannellidae and
Scopelarchidae linkage is much more loose, if valid.
Concerning the possible three main lineages in this order, the
larval evidence is less promising. However, additional larval
evidence regarding developmental sequences, including osteol-
ogy as well as internal morphology, would provide much more
fruitful information for elucidating the phylogeny of this inter-
esting group.
Ocean Research Institlite, University of Tokyo, 1-15-1,
MiNAMiDAi, Nakano-ku, Tokvo 164, Japan.
Gadiformes: Overview
D. M. Cohen
GADIFORMES is a particularly interesting order with which
to work because it encompasses a high degree of diversity
that suggests the existence of several lineages, apparent conver-
gence and reductive trends to trap the unwary, a useful fossil
record that allows a consideration of the distribution in time of
some important taxa and character states, and new suites of
characters based on the study of ELH stages.
Although study of the classification of gadiforms dates from
pre-Linnean times, there is still insufficient properly evaluated
data available to derive a phyletic classification. In fact, there
is not even agreement as to what should be included. Berg (1947)
restricted the order to the muraenolepids, bregmacerotids, mor-
ids, and gadids (including merlucciids) and excluded the mac-
rourids. He noted primitive and advanced characters in his
gadiforms and suggested derivation from primitive fishes. Rosen
and Patterson (1969) revived an expanded Gadiformes dating
at least from the time of Gill, which included not only gadoids
and macrouroids but also ophidioids and zoarcoids, and which
they placed in a supraorder Paracanthopterygii, postulated as
being, "in many ways more primitive than the acanthoptery-
gians" and representing "a spiny-finned radiation more or less
comparable morphologically with that of the Acanthopterygii"
260
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
MERLUCCIUS
BRECMACEROS
EUCLICHTHYS
CADINAE (2)
LOTINAE (2)
EUCLICHTHYS ( u)
MURAENOLEPIS (2)
PHYCINAE 12)
MORIDAE («-5)
MELANONUS ( if- 5)
MERLUCCIUS (21
BRECMACEROS (2)
MELANONUS
HYPURAL RAYS
Fig. 133. Numbers of hypural bones (in parentheses) and fin rays
supported by hypural bones in nine groups of gadiform fishes. Data
from Fahay and Markle (this volume) and original.
TOTAL CAUDAL RAYS
Fig. 132. Total caudal rays in eight groups of gadiform fishes. Data
from Fahay and Markle (this volume) and original.
and including in addition to their gadiforms the polymixoids,
percopsiforms. batrachoids. iophiiforms, and gobiesocoids.
Gosline (1968) analyzed the characters used in defining the ex-
panded Gadiformes and concluded that ophidioids and zoar-
coids are perciform derivatives, while gadoids are widely sep-
arate and probably close to the percopsiforms (Gosline, 1963a).
Marshall and Cohen (1973), whom I follow for present purposes,
restricted the Gadiformes to the gadoids and macruroids but
did not consider the question of relationships. In the following
brief preliminary consideration of the order, I discuss several
characters, mention the groups that I think must be considered,
and outline some of my ideas about the course of evolution in
the gadiforms.
Characters
Several character complexes that require consideration are
discussed below. Others are noted later under groups in which
they occur. Additional relevant information is presented by Fa-
hay and Markle and Dunn and Matarese in subsequent sections
of this volume.
Caudal fin.— Considering the fact that well over half the known
species of gadiform fishes lack the slightest vestige of a caudal
fin, it is a little astonishing how much importance has been
attached to the origin and homologies of the various skeletal
supports and of the fin rays themselves. There is no denying,
however, that when present the gadiform caudal complex is
unique in several respects. Most fish groups may be character-
ized by a set number of branched caudal rays. Furthermore, the
branched rays are generally supported by only hypurals. In gad-
iforms with tail fins, the number of branched caudal rays is
highly variable, as is their skeletal support. Bregmaceros may
have as few as 1 2 branched caudal rays, most of which are
supported by hypurals, while at the upper end of the range, the
lotine Brosmc may have as many as 43 branched rays, which
are supported by hypurals. epurals, and haemal and neural spines.
This high degree of variation in an otherwise conservative an-
atomical complex lends credence to the idea of Boulenger(1902)
and Regan ( 1 903b) that the caudal fin of gadiforms is a structure
newly evolved from an essentially tailless condition such as that
of the macrourids or of some merlucciids. It was partly to test
Regan's hypothesis that Barrington (1937) compared the de-
velopment of the caudal fin of Gadus with that of Pleuwnectes
and concluded that, although the tail of Gadus was unique in
several respects, it could have been derived from an ordinary
homocercal tail that was less specialized than that of Pleuw-
nectes. I agree with Barrington. Barrington commented also on
the presence in gadids of a high number of procurrent caudal
rays, which he interpreted as being far posterior dorsal and anal
rays, so that the functional caudal of a cod is composed of
elements of three fins, dorsal, anal, and caudal proper. This
interpretation has been neither falsified nor verified by the study
of early life history stages. Barrington coined the term pseu-
docaudal for what he took to be this kind of fin. In his lectures
and during conversations with me. Ahlstrom disagreed with
Barrington's explanation and its acceptance by Marshall and
Cohen (1973) because procurrent rays lack pterygiophores. It is
instructive to note in this respect the caudal fin structure of
Muraenolepis (see Fig. 1 43 of Fahay and Markle in this volume),
which has confluent vertical fins and in which the distinctive,
elongate pterygiophores grade into hypurals. It is, in fact, im-
possible to distinguish between the last anal pterygiophore and
the first hypural or parhypural. But see Fahay and Markle later
in this volume.
A variety of controversial interpretations (Gosline, 1963a;
Monod, 1968; Rosen and Patterson, 1969) have been advanced
concerning supposed sequences effusions and deletions of bony
elements in gadiform tails. This particular use of caudal fin
structure in phylogeny has yet to be proven, as few hypotheses
have been verified or falsified.
For purposes of classification within the order, at least four
COHEN: GADIFORMES
261
BREGMACEROTIDAE
EUCLICHTHYS
MERLUCCIUS
10 20 30 40 50
BRANCHED CAUDAL RAYS
Fig. 134. Branched caudal rays in seven groups of gadiform fishes. Data from Fahay and Markle (this volume) and onginal.
caudal fin characters require comment. They are: 1) presence
or absence of a caudal fin; 2) number of hypurals; 3) relationship
between branched caudal rays, hypurals, and procurrent caudal
rays; 4) presence or absence of X- Y bones.
Although vestiges of a caudal fin are sometimes found in a
few macrourid species, it is essentially absent from all of them.
The same is true of the merlucciid genus Lyconus and also
Steindachneria. Loss of a caudal fin has certainly occurred two
times and perhaps more.
The number of hypurals is a useful systematic character. There
are almost always 4 or 5 in morids and Melanonus, and almost
always 2 in gadids, Merliiccius. Bregmaceros, and Muraenolepis:
Euclichthys has 4, nearly fused to 2. 1 consider the lower number
to be an advanced character; the study of developmental series
has verified this interpretation for Raniceps at least (Dunn and
Matarese, this volume). Certainly the loss of hypurals, whether
through deletion or fusion has occurred several times in gadi-
forms.
The evolutionary complexity of the caudal fin in gadiforms
is particularly apparent when considering the numbers of dif-
ferent kinds of caudal fin rays (Figs. 132-134 and Fahay and
Markle, this volume. Table 76). Morids in general have caudal
fins that are small and probably of reduced importance in pro-
pulsion, and which 1 interpret as a derived state; they also have
generally fewer total rays, which Fahay and Markle (this volume)
consider an ancestral state, and unbranched rays that tend to
be short and contribute little to overall caudal fin size; yet,
morids have 4-5 hypurals. Melanonus also has a weakly de-
veloped caudal fin but has 4-5 hypurals and many rays. Gadine
fishes on the other hand, have well-developed caudal fins with
many rays, both branched and unbranched, but have only 2
hypurals. Gadines are in general good swimmers, and one of
the most active of all, Pollachius vtrens, has the most total caudal
fin rays (70 in one specimen) of any gadiform fish. (Sluggish
fishes like the lotines, Brosme and Lota, also have numerous
caudal fin rays but have rounded caudal fins and must swim in
a very different way, probably using the caudal fin as an exten-
sion of the body rather than as an oar.) Although numbers of
different kinds of fin rays may prove useful in taxonomy, the
relationship of branched to unbranched or total caudal fin rays
is variable and has limited apparent value in the present context.
Many gadiform fishes have in their caudal fin skeletons a pair
of bone splints resembling neural and haemal spines. These
structures have been mentioned in the literature as accessory
bones or X and Y bones and have been interpreted as modified
relict pterygiophores or detached neural and haemal spines whose
centra have been lost (Rosen and Patterson, 1969). 1 agree with
Markle (1982) that the absence in any gadiform of X and Y
bones is a derived character.
Dorsal and anal fins.— Gadiform fishes have 1, 2, or 3 external
dorsal fins and 1 or 2 external anal fins. The number, size, and
location of these fins have been used for hundreds of years to
characterize groups of species. Prior to the recognition of Mor-
idaeasa distinct family (Svetovidov, 1937), convergence in this
character was not recognized; most ichthyologists lumped
gadids and morids with similar fin patterns.
Svetovidov ( 1948) assumed on functional grounds that a sin-
gle dorsal and single anal is the primitive condition and arranged
the gadid genera in a transition series based on increasing num-
ber of fins and the distance of their separation from each other.
His hypothesis is supported by the presence in all gadiforms of
a single, continuous, postanal series of pterygiophores, present
even over areas that lack fin rays. Complete or partial division
of the exterior fin has occurred several times, for example in
the gadines, Euclichthys. Merluccius. and in the morid genera
Mora, Halargyreus. Lepidion, Laemonema. and Tripterophycis.
Although only a few gadiforms have a single dorsal fin, the
condition has a broad taxonomic distribution; examples are the
gadid Brosme. the merlucciid Lyconus. Melanonus, and the ma-
crouroidine rattails. Nearly all gadiforms have 2 or 3 dorsals,
but even in those with 3, there are only two series of pteryg-
iophores. From fewer to more dorsals would seem to be a rea-
sonable transition series. But it certainly has occurred more than
once, even within Gadidae, as Markle (1982) has demonstrated.
Pectoral radials. — Mosx gadiforms have five pectoral radials.
Muraenolepis has more; Bregmaceros has fewer; both are in-
terpreted as derived conditions.
First neural spine. — Many gadiforms have the first neural spine
closely adpressed to the occipital crest. I take this as a derived
character. Muraenolepis has a free spine, but it is modified by
the presence of a prominent wing-shaped enlargement extending
on either side of the occipital crest.
Olfactory lobes. — In his classical monograph on the Gadidae,
Svetovidov (1948) discussed the position of the olfactory lobes
262
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 135. Dorsal view of cranium in three genera of gadiform fishes; left. Rhinocephalus planiceps: center, Palaeogadus intergerinus; right,
Merluccius merluccius. From Fedotov ( 1 976).
of the brain and used their advanced position, adjacent to the
nasal capsule, as his primary character for defining the Gadi-
formes. This is a derived character, which has been found also
in cyprinids, galaxiids, and mormyrids. Svetovidov noted that
the olfactory lobe is located in an intermediate position in the
gadid Raniceps. A posterior location of the lobe was subse-
quently recorded in Melanonus and several macrourids and an
intermediate location in merlucciids, Steindachneria, the gadid
Raniceps, and two macrourids (Marshall and Cohen, 1973).
Svetovidov ( 1 969) pointed out the size dependent nature of this
character, especially in Merluccius (which I have verified in M.
bilinearis and M. productus). Further investigation is required,
especially in species that mature at small sizes.
V-shaped crest on skull.— As long ago as 1903b Regan noted
the shared presence in Merluccius and Macruronus of prominent
V-shaped ridges on the frontals, which converge on the supra-
occipital crest. These structures have subsequently been found
in the extinct genera Rhinocephalus and Palaeogadus (Fig. 135)
as well as in some fossil percopsiforms (Rosen and Patterson,
1969) and are present in varying degrees in Lyconus and Stein-
dachneria.
Groups and Their Relationships
In this section I briefly discuss those taxonomic units that I
think require consideration and explain as best possible the
reasons for their placement on Fig. 1 36.
"Protocodus" is an unnamed species' from the Paleocene of
Greenland (discussed by Rosen and Patterson, 1969 and Fe-
' The name "Protocodus" is used as a designation of convenience
and does not have formal, nomenclatural significance.
dotov, 1 976; I too have examined it), which is the oldest known
non-otolith gadiform. It has a number of characters that may be
interpreted as primitive for the group, including five, slender,
well-separated hypurals, X-Y bones, numerous procurrent rays,
and a V-shaped ridge on the frontals. It has a dorsal and anal
fin configuration much like that of Merluccius (Rosen and Pat-
terson, 1969).
Muraenolepis is a highly distinctive genus with four or more
species. It has such primitive characters as a single anal and
long-based second dorsal fin, a dermal basibranchial plate (Ro-
sen and Patterson, 1969), the similarity of the lower hypurals
to pterygiophores and to caudal fin elements, and a free first
neural spine. Derived characters include 12-14 pectoral radials,
a single epural, first dorsal fin a single-rayed anteriorly placed
filament, vertical fins confluent around the tail, an oblique pat-
tern of squamation, and modifications of the first neural spine.
Muraenolepis is not obviously related to any other gadiform
and appears to represent an ancient lineage.
Bregmaceros is another distinctive genus with no obvious
close relatives. Like Muraenolepis it retains a dermal basi-
branchial plate, but this is a primitive character, as is possession
of a uroneural and a set of X-Y bones in the tail. Derived
characters include the conjunction of the first neural spine with
the occipital crest, a large consolidated hypural plate supporting
many branched rays, a unique lateral line system, only two
pectoral radials, and a long dorsal ray on top of the head. The
tropical pelagic habitat of these fishes is also different from that
of any other gadiform. If fusion of the first neural spine with
the occipital crest has occurred only a single time, then Breg-
maceros must have originated after Rhinocephalus.
Rhinocephalus is an Eocene fossil, the skull of which has been
described in some detail and compared with other gadiforms
by Rosen and Patterson ( 1 969). They mention and illustrate a
COHEN: GADIFORMES
263
RECENT.
PLEISTOCENE.
PLIOCENE.
MIOCENE.
OLIGOCENE.
EOCENE.
PALEOCENE.
"PROTOCODUS"
Fig. 136. Phylogenetic bush showing hypothetical inter-relationships among gadirorm fishes. Beginning of soHd Unes based on fossils, not
including otoliths or scales.
V-shaped indge on the frontals and also lateral flanges on the
rear of the skull that characterize gadines and at least some
morids. They write, "The skull roof of Rhinocephalns shows
many features common to morids, merlucciids, gadids. and
macrourids . . . ." In addition, the first neural spine is free from
the supraoccipital crest.
Eucltchthys (Fig. 137), represented by a single South Austra-
lian and New Zealand species, was incorrectly placed in Moridae
but removed by Svetovidov (1969), who pointed out some sim-
ilarities to Macrouridae. Enclichlhys can not be placed in any
currently recognized family. It has a free first neural spine, which
may indicate an origin prior to Palaeogadus. lacks an otophysic
connection, has four hypurals nearly fused to two, and in two
specimens has only one of the X-Y bones. As in morids, which
are more specialized than macrourids and could not have given
rise to them, Eitclichthys has an asymmetrical, rather reduced
caudal fin. Perhaps this curious fish is a modem representative
of a macrourid progenitor.
Macrouroidinae is represented by two small genera and has
been treated both as a subfamily of Macrouridae (Marshall,
1973) and a separate family (Okamura, 1970a). It has single
dorsal and anal fins and a number of distinctive features in the
head skeleton and may represent the most primitive tail-less
macruroid.
Macrourinae-Trachyrincinae, which may well constitute two
quite separate groups, has 20-25 genera and contains more than
half of all gadiform species (Okamura, 1970a; Marshall, 1973).
The caudal fin is absent in most, vestigial in a few; the first
neural spine is free, and there is no V-shaped ridge. Eggs of the
few species for which information is available have a distinctive
hexagonal pattern; many species have light organs.
Bathygadinae, with two genera, differs from other macrourids
in having a large, terminal mouth, dorsal rays longer than anal
ones, and in a variety of other ways summarized by Okamura
(1970a), who interprets most of the bathygadine characters as
primitive ones. Differences in functional morphology between
bathygadines as pelagic feeders and macrourines as benthic to
benthopelagic feeders have been described by McLellan (1977).
Melanonus has two meso-to-bathypelagic species formerly
placed in Moridae, where they do not belong as they lack an
otophysic connection, have a single dorsal fin, and have lost the
X-Y bones. Otherwise, they seem similar to Moridae. The first
neural spine is joined to the occipital crest, suggesting an origin
after Rhinocephalus. A separate family was proposed by Mar-
shall (1965).
Moridae consists of 12-15 genera, some highly diverse, and
all characterized by possession of an otophysic connection, 4 or
5 hypurals, X-Y bones, a joined first neural spine, and distinctive
otoliths; many species have light organs. Morids probably di-
verged from the main Rhinocephalus-Palaeogadus-Merluccius
264
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Fig. 137. Euclichlhys polynemus, holotype. From McCulloch (1926).
line after fusion of the neural spine and at least some of their
evolution is in parallel with the gadids.
Palaeogadits is a well-known Eocene fossil genus in which the
V-shaped crest has been retained, but specializations include a
joined first neural spine and only two hypurals. It is, in fact,
very similar to modem Merluccius. DaniPchenko (1950), who
reviewed Palaeogadus. believed that it gave rise independently
to Lotinae and Gadinae as well as to Merluccius.
Phycinae, as recently modified by Markle (1982), is presently
included in the family Gadidae. Fahay and Markle (this volume)
would like to escort it out. An early Oligocene fossil genus,
Eophycis (Jerzmanska, 1968) has been suggested as a precursor
of Phycts and Urophycis, and probably arose independently of
other gadid subfamilies, which supports Fahay and Markle's
position.
Lotinae is a gadid subfamily that I mainly leave to Fahay and
Markle and Dunn and Matarese. I note, however, Mujib's ( 1 967)
conclusion based on cranial osteology that Lotinae could have
arisen from Merlucciinae. Lotines have no V-shaped crest but
retain X-Y bones. Hypurals are two, the first neural spine is
joined to the occipital crest, and there are more branched rays
than in any other gadid.
Gadinae has about a dozen genera, all of which have three
external dorsal and two external anal fins and a large caudal,
even though there are only two hypurals. Derived characters
include fused frontals, absence of X-Y bones, and a joined neu-
ral spine; Fahay and Markle and Dunn and Matarese (this vol-
ume) give more.
Merluccius, with about a dozen closely related species (Inada,
1981b), has been treated as the type of a separate family or as
a subfamily of Gadidae. Primitive characters include a V-shaped
ridge and X-Y bones. Advanced ones are the joined first neural
spine and the reduced number of hypurals. Merluccius appears
to be the modem representative of a lineage commencing with
"Protocodus" and extending through Rhinocephalus and Pa-
laeogadus. which it closely resembles (Rosen and Patterson,
1969).
Macruronus, which has three nominal species found in tem-
perate waters of the southern hemisphere, is basically a Mer-
luccius wiih a much reduced caudal fin. I mention it here because
it has been referred incorrectly to Macrouridae and considered
by some to be a link between that family and Merluccius.
Lyconus. with several pelagic oceanic species, is probably re-
lated to Merluccius. It lacks a caudal fin and has a single dorsal
fin.
Steindachneria, is a monotypic tropical western Atlantic ge-
nus with luminescent organs, a wide separation between the
anus and urogenital openings, and no caudal fin. It has been
placed in Macrouridae and also considered a separate family
(Marshall and Cohen, 1973). It may be closer to Merluccius
than to any other known gadiform.
Classification
How best to classify gadiforms for working purposes in a way
that approximates their possible phylogenetic relationships is
diflicult because the existence of fossils, which appears to help
indicate lineages, creates logical traps for the classifier. The fol-
lowing arrangement, unfortunately based on gaps for some groups
and on a continuum for others, is an interim suggestion for
further testing.
Euclichthys is accorded family status for the first time because
it can not be placed in any gadiform family. Gadidae is restricted
to the gadines, and Lotidae and Phycidae are recognized at the
full family level (family group names for the latter two date at
least from Goode and Bean, 1 896), because available evidence
indicates an independent origin from Palaeogadus for each of
the three groups. If merlucciids were reduced to subfamily rank
and placed with gadines, lotines, and phycines in a more inclu-
sive family Gadidae, then consistency would require the inclu-
sion of at least two other well-defined apparent derivatives of
the Rhinocephalus- Palaeogadus- Merluccius stem, Moridae and
Melanonidae. In the present instance I believe that splitting is
more useful than lumping.
Suborder Muraenolepoidei
Family Muraenolepidae
Suborder Bregmacerotoidei
Family Bregmacerotidae
Suborder Macrouroidei
Family Euclichthyidae
Family Macrouridae
Subfamily Macrouroididae
Subfamily Trachyrincinae
Subfamily Macrourinae
Subfamily Bathygadinae
Suborder Gadoidei
COHEN: GADIFORMES
265
Family Merlucciidae
Subfamily Merlucciinae (including "Protocodus," Rhi-
nocephalus. Palaeogadus, Merluccius. Macruronus,
and Lyconus)
Subfamily Steindachneriinae
Family Gadidae
Family Lotidae
Family Phycidae
Family Moridae
Family Melanonidae
Life Sciences Division. Los Angeles County Museum of
Natural History, 900 Exposition Boulevard, Los
Angeles, California 90007.
Gadiformes: Development and Relationships
M. P. Fahay and D. F. Markle
AS treated herein, the Gadiformes includes about 63 genera
and 400+ species (Nelson, 1976) divided into eight fam-
ilies (Gosline, 1968; Marshall and Cohen, 1973); (but see Cohen,
this volume). They are primarily marine with familial distri-
bution "centers" as follows: Muraenolepididae— high latitudes,
southern hemisphere; Bregmacerotidae — tropical and sub-
tropical, world-wide; Melanonidae— tropical and sub-tropical,
world-wide; Moridae— world-wide; Macrouridae— deep sea,
world-wide; Steindachneriidae— tropical W. Atlantic; Merlucci-
idae—mid-latitudes, both hemispheres; and Gadidae— high lat-
itudes, northern hemisphere with minor freshwater and south-
em hemisphere components.
Meristic characters of genera within each family are presented
in Table 72 (except that macrourid characters will be found in
Table 75). Gadiforms characteristically have relatively high ver-
tebral counts, with caudal centra outnumbering precaudal cen-
tra, usually by a wide margin. The first two centra lack ribs and
parapophyses. Vertical fins have numerous rays and long bases,
with posterior dorsal and anal rays separate from caudal fin rays
except in Miiraenolepis and macruronines. Pectoral fins are typ-
ically high on the body and pelvic fins typically thoracic or
jugular in position. Mental barbels are found in many genera
and mouth position ranges from terminal to inferior.
Present State of KnowLedge and
Characters of Early Life
History Stages
Literature on gadiform eggs and larvae is heavily weighted
towards gadids and merlucciids. within which the commercially
important gadines and Merluccius have received most attention.
Gadine larvae were among the first marine fish larvae to be
described. In fact. G. O. Sar's discovery, early in the 1860's,
that cod eggs and larvae were pelagic, helped initiate fisheries-
oriented ichthyoplankton surveys. In addition to their com-
mercial importance, gadines and Merluccius are found in shelf
waters where their early stages are more accessible than those
of other gadiforms which are largely residents of slope and oceanic
waters.
Published descriptions of gadiform early life history stages
are listed in Table 73. We especially note the seminal work on
young gadids done by Johannes Schmidt in the early 1900's.
Although he stressed pigment patterns over other develop-
mental features. Schmidt was one of the first to look at several
species in a systematic fashion.
In the following review, we summarize gadiform characters
in brief family synopses as well as through a limited survey of
the ontogeny of selected characters. Our purposes are, respec-
tively, to point out what appear to be easily observed diagnostic
early life history characters and to contribute to discussions of
gadiform phylogeny.
Gadiformes.— The gut of gadiform larvae coils early in ontogeny
and combined with a tapering postanal region and rounded
head, contributes to an overall tadpole-like appearance. These
features are, in part, a reflection of vertebral and vertical fin ray
elements (Table 72) and are not diagnostic. Although it has not
been documented in all families and is not always easily ob-
served, yolk-sac and first-feeding gadiform larvae have an anus
that exits laterally through the finfold rather than medially as
is usual in teleost larvae. Some secondary caudal rays develop
before some primary in forms with a caudal fin.
In Table 74 we summarize some developmental features of
each family. A rather widespread trend is for the pelvic fin to
be the earliest forming fin. There does not seem to be any char-
acter unique or diagnostic for young gadiforms. The features of
body shape, anus morphology and pelvic fin development in
combination with specific familial characters appear to be the
most useful for initial identification. Transformation is gradual
and direct with no striking changes in ontogeny.
Muraenolepididae. — \ single planktonic juvenile (see discus-
sion of planktonic juveniles below) of Muraenolepis sp. is shown
in Fig. 138 A. The distinctive first dorsal fin, composed of one
or two rays, the confluent vertical fins, meristic characters (Ta-
bles 72 and 76), chin barbel, restricted gill opening and capture
locality (53°48.7'S, 38°18.7'W) preclude all other teleosts and
agree with characters described for Muraenolepis (Svetovidov,
1948). The lateral premaxillary spines (Fig. 138A) were not
shown in a schematic illustration of an early Muraenolepis (North
and White. 1982) or in larvae described by Efremenko (1983b)
and are not reported for adults. It is possible that they are not
found in larvae of all species of Muraenolepis, but for present
purposes we consider them a unique and diagnostic larval spe-
cialization of the family.
266
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 72. Meristic Characters in Gadiformes. (See Table 75 for characters of the Macrouridae.) Characters of the caudal fin are contained
in Table 76. "Number of species" includes number of nominal species followed by number surveyed for meristic characters. Primary sources of
data: Gunther, 1887; Goode and Bean, 1896; Ehrenbaum, 1905-1909; Thompson, 1916; Norman, 1930; D'Ancona, 1933a; Parr, 1946; Jensen,
1948; Svetovidov, 1948; Koefoed, 1953; Andriyashev, 1954; Rass, 1954; Smith, 1961; Scott, 1962; Lmdberg and Legeza, 1969; Leim and Scott!
1966; Templeman, 1968; Fitch and Barker, 1972; Miller and Lea, 1972; Hart, 1973; Inada and Nakamura, 1975; Brownell, 1979; Cohen, 1979;
Cohen and Russo, 1979; Inada, 1981a; Inada, 1981b; Matarese et al., 1981; Yabe et al., 1981; Demir, 1982; Markle, 1982; Fahay, 1983; Paulin,
1983.
Family
genus
Number
of
species
Prc-
caudal
Vertebrae
Caudal
Total
Fin rays
D,
Dj
D,
D total
A,
A.
A total
Pelvic
Pectoral
Muraenolepididae
Muraenolepis 4 + /1 20-21 46-49 67-69
Bregmacerotidae
Bregmaceros 9 + /7 13-14 32-42 43-59
127-141
34-65
129-142 98-112 98-112 4 37-38
35-66 42-69 42-69 5-7' 16-21
or
15-21 27-31
Melanonidae
Melanonus
111
13
47
58-62
72-78
or
5-8
59-70
72-78
50-58
50-58
5-7
10-16
Fam. Incertae sedis
Euclichthys
?/l
15
55
70
15
74
89
15
77
92
5
20
Moridae
Antimora
2/2
24-25
33-35
57-61
4-7
48-56
54-60
36-49
36-49
5-7
17-25
Auchenoceros
1/1
11
37
46-49
1-2
13
51
65 (+10)
62-82
62-82
2'
23
(+10)
(holo-
type)
Brosmiculus
1/1
17
33-34
50-51
58 or
10
53-56
58
63-66
56-62
56-62
5
—
Eretmophorus
1/1
—
—
—
4-5
66-77
70-82
64-73
64-73
5
22
Gadella
1/1
—
—
50
9-12
55-64
64-74
56-67
56-67
6-7
20-25
Halargyreus
1/1
—
30-35
51-58
6-8
47-60
53-67
17-26
21
-29
39-53
5
17-20
Laemonema
147/14
15-17
42-45
50-63
5-6
48-75
53-80
45-72
45-72
(1)2(3)
15-26
Lepidion
l->/l
17-18
42-45
54-63
4-7
49-62
54-68
40-54
40-54
(5) 6-8
17-23
Loiella
6?/6
13-15
27-34
41-50
5-8
46-69
51-73
42-61
42-61
7-9
19-26
Mkroteptdium
1/1
11-12
33-34
44-46
7-9
39-42
46-51
39-42
39-42
2
19-22
Mora
1/1
15
35
50-54
7-11
42-53
49-60
16-22
15
-22
30-44
5-6
18-25
Physiculus
127/12
12-16
34-42
48-59
7-11
44-71
53-79
43-79
43-79
(3) 5-7
20-28
Pseudophyas
3/3
—
—
42-51
8-14
40-63
51-67
39-68
39-68
5-6
19-27
Rhynchogadus
1/1
_
_
_
5
50-56
55-61
44-49
44-49
7
20-22
Salilota
1/1
—
—
—
9-11
56
65-67
57
57
—
_
Svelovidovia
2/2
14-15
43
57-58
5-7
57-63
62-70
56-62
56-62
10-11
17-18
Tripterophycis
7/2
-
-
67-72
4-7
12-17
29-
-39
48-58
95-112
95-112
5
15-20
Gadidae (Lotinae)
Brosme
1/1
19-21
44-46
63-66
85-108
85-108
62-77
62-77
5
22-24
Lola
1/1
23-26
37-39
59-66
9-16
65-93
75-108
63-85
63-85
6-8
18-21
Molva
3/3
25-36
37-48
63-84
10-16
61-85
74-98
57-82
57-82
6-7
18-21
Gadidae (Phycinae)
Ciliata
2/2
(12)
13-15
30-34
44-48
1 +
45-55
46-56 +
40-46
40-46
7'
15-17
Enchelyopus
1/1
15-17
38-39
49-55
1 +
45-52
46-53 +
34-49
34-49
5-6 (7)'
15-19
Gaidropsarus
14/14
13-17
32-36
44-53
1 +
45-70
46-71 +
38-60
38-60
5-9'
14-24
Phycis
3/3
15-16
or
18-19
28-29
or
32-35
8-11
54-65
63-70
47-65
47-65
(2)3'
15-19
Raniceps
1/1
11
33-34
44-45
3
61-67
64-70
55-61
55-61
6
21-22
Vrophycis
7/7
13-17
30-37
44-52
(56, 57)
8-13
43-68
53-78
40-58
40-58
2-
15-18
Gadidae (Gadinae)
Arclogadus
2/2
19-22
36-40
54-62
10-16
15-24
19-
-25
17-24
18-
-25
6-7
17-23
Boreogadus
27/2
18-20
35-39
49-58
9-16
12-19
16-
-25
49-55
14-23
18-
-24
39-44
6
18-21
Eleginus
2/2
21-24
37-41
55-64
11-16
14-24
18-
-24
19-27
18-
-26
6
18-22
Gadiculus
1/1
—
—
39-43
9-12
11-17
15-
-18
12-19
15-
-19
6
14-15
Gadus
77/3
18-22
31-37
49-58
10-17
11-24
10-
-22
45-62
16-27
12-
-25
35-48
6-7
19-22
FAHAY AND MARKLE: GADIFORMES
267
Table 72. Continued.
Mela nogra mm us
Merlangius
Microgadus
Micromesislius
Pollachius
Theragra
Trisoplerus
Merlucciidae
Merluccius
Macruronus
Lyconus
Steindachneriidae
Steindachneria
Number
of
species
Venebrae
Fin rays
Family
fienus
Prc-
caudal
Caudal
Tolal
D,
Dj
D,
D lolal
A,
A;
A total
Pelvic
Pectoral
1/1
2/1
2/2
2/2
2/2
2/1
3/3
1/1
19-21 33-36 52-57 14-18
23(?) - 53-57 12-17
17-22 34-38 53-60 9-15
24-26 30-33 54-60 11-14
20-23 32 52-56 11-15
18-20 31-34 48-52 10-14
- - 44-55 11-16
12/12 21-29 24-31
3/3 20-21' 58-60'
2/1
48-58
78-81'
13
1,7-12
8-11
90 +
8-12
19-26
18-25
15-21
10-15
16-24
12-19
16-28
34-45
105-120
123 +
19-24
19-22
16-24
21-27
15-24
14-23
16-27
56-67
21-28
28-38
(12)
18-29
33-41
23-34
15-24
25-36
35-46
86-105
90+ 65 +
20-25
20-25
16-28
22-30
16-24
15-23
17-30
45-53
10-11 + 113
86-105
65 +
6-7
6
6-7
6
6
6-7
6
(6) 7 (8)
9-10
10
19-21
19-20
(16)
18-19
18-23
17-20
18-21
(13)
17-19
12-18
14-18
13
14-15
' Four rays in larvae.
- Three rays in larvae.
^ n = 2 (A/, novaezelandiae).
Bregmacerotidae. — Larval and juvenile bregmacerotids appear
distinctive in the early acquisition of a cephalic dorsal fin ray.
Larvae have been described (Table 7 3 ), but eggs are undescribed.
Characters are reviewed in this volume by Houde.
Melanonidae. —Eggs, larvae and young stages have not been
previously described for melanonids (Cohen, 1973). Early stages,
however, are moderately abundant in some oceanic collections.
In the smallest specimens seen by us (ca. 15 mm SL) the fins
are all formed and they have the general body shape of adults
(Fig. 1 38B). Notable features of this stage are the small eye, dark
peritoneum and distinctive caudal fin.
Monrfae.— Considering the diversity of the family, very little is
known of the early life history stages of morids. Eggs with oil
globules have been described for Physiculus dalwigkii (De
Gaetani, 1928), Mora mora (D'Ancona, 1933a), Physiculus ca-
pensis (Brownell, 1979), Salilota australis (de Ciechomski and
Booman, 1981) and Laemonema longipes (Kuroda et al., 1 982).
Pelagic juveniles of some morids have not yet been related
to adult forms and have been placed in three genera, Rhyncho-
gadus Tortonese, 1948; Svetovidovia Cohen, 1973 and Eret-
mophorus Giglioli, 1889. One of these forms, S. vitellius (Koe-
foed, 1953), is shown in Fig. I38C, D. This form appears to be
the juvenile stage of Laemonema. In our largest specimen, 55
mm SL (MCZ 59773), the pelvic fin has two rays plus two or
three remnants. This is a reduction from a count of 9-1 1 in
smaller specimens. To date it has not been possible to assign
this form to a known adult. A second type of Svetovidovia is
shown in Fig. I38E. D'Ancona (1933a) suspected that Eret-
mophorus kleinenbergi was the young of Lepidion lepidion but
Cohen (1973) apparently was not convinced of the relationship.
Finally, Rhynchogadus Tortonese, 1948 (= Hypsirhynchus) is a
pelagic form referrable to no known adults and may also rep-
resent an early stage of a species whose adult form is known
under another name.
The early stages of morids appear stocky anteriorly, with well
developed to voluminous pelvic fins, frequently with more in-
ferior than superior procurrent caudal fin rays, and relatively
voluminous posterior sections of dorsal and anal fins (see Figs.
138C-E, 139A-F). Earliest stages may be difficult to separate
from some merlucciids and gadines.
Macrouhdae.— There is a moderate amount of early life history
information available on macrourids but considering that the
family contains over a third of all extant gadiform species (Nel-
son, 1976), a great deal remains unknown. Eggs have been de-
scribed by Gilchrist (1904), Sanzo (1933a), de Ciechomski and
Booman (1981) and Grigorev and Serebryakov (1981). All de-
scribed macrourid eggs range from about I to 4 mm (Marshall
and Iwamoto, 1973). Most are less than 2 mm, have a single
oil globule and characteristic honey-comb ornamentation on
the chorion (see Boehlert, this volume).
Larvae and pelagic juveniles have been infrequently described
and only for macrourines (Table 73). Early ontogenetic stages
of trachyrhynchine and macrouroidine macrourids are still not
known though Johnsen (1927) illustrated and discussed meta-
morphosed Trachyrhyncfius juveniles.
Only one pelagic juvenile bathygadine is known (Fig. 1408).
The specimen, tentatively referred to Gadomus. can be recog-
nized by its long second dorsal (relative to anal) fin rays, short
interspace between dorsal fins, moderate-sized barbel and fine
jaw teeth. Other important characters of this specimen are its
laterally placed thoracic pelvic fins and paired preopercular skin
flaps. The general appearance of young bathygadines approaches
that of morids, with the lack of caudal fin and presence of
pedunculate pectoral fins the obvious differences.
Numerous macrourine larvae have been described. The spec-
imen illustrated as Fig. 140C appears identical to Johnsen's
(1927) "AH 1" macrourid larva while Fig. MOD is similar in
appearance to Merrett's (1978) Coryphaenoides rupestris. In both
cases meristic characters agree with Coryphaenoides (sensu lato),
but we are unable to provide further identification at this time.
A number of more elongate types are also known. A specimen
belonging to either Cetonurus or Nezumia is shown in Fig. 1 4 1 A.
A similar specimen, also with seven branchiostegals, is shown
in Fig. 14 IB; its meristic characters, however, do not permit
268
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 73. Published Descriptions of Early Life History Stages in Genera within Eight Gadiform Families.
Family
genus
Family
genus
Muraenolepididae
Muraenolepis
Moridae
Auchenoceros
Erelmophorus'
Gadella
Mora
Gaidropsarus
Physiculus
Rhynchogadus'
Salilota
Svetovidovia
Melanonidae
Melanonus
Bregmacerotidae
Bregmaceros
Gadidae (Lotinae)
Brosme
Lota
Molva
Gadidae (Phycinae)
Ciliata
Enchelyopus
North and White, 1982
Efremenko, 1983b
Robertson, 1975a
Crossland. 198P
Mazzarelli, 1917
Lo Bianco, 1911
Sparta, 1928
Lo Bianco, 1911
DeGaetani, 1926
D'Ancona, 1933a
Lo Bianco, 1911
Mancuso, 1926
DeGaetani, 1928
D'Ancona, 1933a
Pertseva-Ostroumova and Rass, 1973-
Brownell, 1979
Lo Bianco, 1911
Cipna, 1927
D'Ancona, 1933a
Weiss, 1975
de Ciechomski and Booman, 1 98 1
Koefoed, 1953
Fahay, 1983^
None
Munro, 1950
Clancey, 1956
D'Ancona and Cavinato, 1965
Aboussouan, 1968c
Pertseva-Ostroumova and Rass, 1973
Belyanina, 1974
Houde, 1981
Mcintosh, 1893
Schmidt, 1905b
Ehrenbaum, 1905-1909
Rass, 1949
Ehrenbaum, 1905-1909
Meshkov, 1967
Jude, 1982b
Mcintosh, 1893
Mcintosh and Masterman, 1897
Heincke and Ehrenbaum, 1900
Schmidt, 1906b, 1907b
Ehrenbaum, 1905-1909
D'Ancona, 1933a
Russell, 1976
Ehrenbaum, 1905-1909
Dando, 1975
Russell, 1976
Agassiz, 1882
Agassiz and Whitman, 1885
Brook, 1890
Ehrenbaum and Strodtman, 1904
Ehrenbaum, 1905-1909
Dannevig, 1919
Colton and Marak, 1969
Russell, 1976
Phycis
Raniceps
Urophycis
Gadidae (Gadinae)
Arctogadus
Boreogadus
Eleginus
Gadiculus
Gadus
Melanogrammus
Merlangius
Microgadus
Micromesistius
Roule and Angel, 1930
D'Ancona, 1933a
Vodyanitsky and Kazanova, 1954
Fives. 1970b
Schmidt, 1905a, 1906a
Ehrenbaum, 1905-1909
Russell, 1976
Bini, 1971
Dekhnik, 1973
Brownell, 1979
Demir, 1982
Markle, 1982
Facciola, 1882
Emery, 1886
Manon, 1894b
D'Ancona, 1933a
Russell, 1976
Heincke and Ehrenbaum, 1900
Schmidt, 1907b
Ehrenbaum, 1905-1909
Kennedy and Fitzmaurice, 1969
Russell, 1976
Aggassiz, 1882
Aggassiz and Whitman, 1885
Hildebrand and Cable, 1938
Bigelow and Schroeder, 1953
Miller and Marak, 1959
Barans and Barans, 1972
Serebryakov, 1978
Zvyagina, 1961
Schmidt, 1905a, 1906a
Rass, 1949
Kuz-min-Karovaev, 1930
Khaldinova, 1936
Ponomareva, 1949
Rass, 1949
Mukhacheva, 1957
Aronovich et al., 1975
Dunn and Vinter, 1984
Schmidt, 1906a
Roule and Angel, 1930
Heincke and Ehrenbaum, 1900
Masterman, 1901
Schmidt, 1905a, 1906a
Dannevig, 1919
Uchidaet al., 1958
Mukhacheva and Zvyagina, I960
Russell, 1976
Matarese et al., 1981
Mcintosh and Pnnce, 1890
Heincke and Ehrenbaum, 1900
Schmidt, 1905a, 1906a
Dannevig, 1919
Russell, 1976
Heincke and Ehrenbaum, 1900
Schmidt, 1905a, 1906a
Ehrenbaum, 1905-1909
D'Ancona, 1933a
Dekhnik, 1973
Russell, 1976
Booth, 1967
Matarese et al., 1981
Schmidt, 1905a, 1906a
D'Ancona, 1933a
Seaton and Bailey, 1971
FAHAY AND MARKLE: GADIFORMES
269
Table 73. Continued.
Family
genu'.
Family
senui
Source
Steindachneriidae
Steindachneria
None''
Macrouridae
Ateleobrachium^
Gilbert and Burke. 1912
Coetorhynchus
Sanzo, 1933a
de Ciechomski and Booman, 1981
Gilchnst, 1905
Coryphaenoides
Johnsen, 1921
Merrett, 1978
Stein, 1980b
Grigorev and Serebryakov, 1 98 1
Hymenocephalus
Sanzo, 1933a
Krohnius^
Costa, 1869
Smitt, 1895
Rouleand Angel, 1930?
Sanzo, 1933a
"Macrouridae"
Ehrcnbaum, 1905-1909^
Murray and Hjort, 1912
Johnsen, 1927
Evseenko. 1982b
Macrourus
Yanulov, 1962
de Ciechomski and Booman, 1981
Efremenko, 1983a
Malacocephalus
Marshall, 1964
Mesobtus
Hubbs and Iwamoto, 1977
Odontoinacrurus
Maul and Koefoed, 1950
Maul, 1951
Koefoed, 1953
Marshall, 1964
ISphagehra nchusl^
Backus etal., 1965
Trachyrhynch us
Johnsen, 1927
Pollachius
Theragra
Tnsoplerus
Merlucciidae
Lyconus
Macruronus
Merluccius
Weiss, 1974
Russell, 1976
Coombs and Hiby, 1979
de Ciechomski and Booman, 1981
Lisovenko et al., 1982
Mcintosh, 1893
Gorbunova, 1954
Matarese et al., 1981
Mcintosh, 1893
Schmidt, 1905a, 1906a
Ehrenbaum. 1905-1909
D'Ancona, 1933a
Rass, 1949
Russell, 1976
None
None
Aggassiz and Whitman, 1885
Raffaelle, 1888
Schmidt, 1907a
Ehrenbaum, 1905-1909
Kuntzand Radcliffe, 1917
D'Ancona, 1933a
Ahlstrom and Counts, 1955
Miller, 1958
Fischer, 1959
Marak, 1967
Sauskan and Serebryakov, 1968
Santander and de Castillo, 1969
Colton and Marak, 1969
de Ciechomski and Weiss, 1974
Russell, 1976
Brownell, 1979
Markleetal., 1980
Fahay, 1983
' No adull specimens.
' May refer to Svetovidovia. a larval stage name,
' Illustration only
* No published descnptions Mead ( 1 963) reports collection of 1 5 larvae, 9.0 to 66.0 mm SL (MCZ 43083).
' Name applied to lar\'al stage Referred to Coryphaenotdes acrolepts by Johnsen (1927).
^ Name applied to larval stage Probably Nezunita (Marshall and Iwamoto. 1973).
' Ehrenbaum's "Macrtindae" plate (hg. 108) illustrates a Mauroltcus egg, a Lophius larva, a percifonn larva possibly referrabte to Carangidae or Serranidae and a 92-mm macround larva
resembling Krohntti^
" Possibly referrable to Atlantic specimen of Mesobnts (see Hubbs and Iwamoto, 1977).
identification by process of elimination anci its identity must
await further study. One of the most elongate of the known
macrouind larvae is that of Odontomacrurus murrayi (Maul and
Koefoed, 1 950). Among these elongate types there is a tendency
for caudal spotting, either as supranal melanophores(Fig. 141 A)
or as midlateral spots or bars (Fig. 14 IB). With development,
there is a marked change in mouth orientation from oblique to
almost horizontal in the elongate Mesohius herryi (Hubbs and
Iwamoto, 1977) (Fig. 142C) and in Coryphaenoides (Stein,
1980b).
Known macround larvae can be characterized by their mod-
erate (Gadomus) to very elongate tail, lack of caudal fin and
moderate to very elongate pectoral fin peduncle. Some adult
diagnostic characters (Table 75), such as numbers of branchio-
stegal rays and retia mirabilia are present early and are crucial
to identification (Merrett, 1978; Stein, 1980b), while others,
such as dorsal spine serrations, develop late and cannot be used
(Merrett, pers. comm., and unpublished observations). Addi-
tional characters (the interspace between dorsal fins, anterior
extent of anal fin origin, the position of fin origins relative to
centra or myomeres, the relative size and shape of pectoral fin
peduncles and larval pigmentation) are not known or not re-
ported, but appear to offer promise in characterizing groups of
larval macrourids (see Table 75 and Figs. 140, 141A-B and
I42C).
Steindachneriidae.— Steindachneria was aligned with the ma-
crourids in early works (Jordan and Evermann, 1896-1900),
with merlucciids by Norman ( 1 966), Marshall ( 1 966b) and Nel-
son ( 1976), and as a separate family (Marshall and Cohen, 1 973).
Eggs are not known and larvae have not previously been de-
scribed although Mead (1963) mentions specimens, 9.0 to 66.0
mm SL (MCZ 43083). An early planktonic juvenile was avail-
able and is illustrated in Fig. 142B-C. Noteworthy features are
the distinctive striated photogenic organs on the ventral surface
of the gut (Cohen, 1964a), genital papilla and orifice separated
270
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 74. Developmental Characters in Gadiform Families and Gadid and Merlucciid Subfamilies.
Muraenolepididae
Bregmacerotidae
Mclanonidae
Mondae
Gadidae
Lolinae
Phycinae
# dorsal fins (exter-
2
1 single ray on head
1 (deeply
2(3)
1-2
2 (first some-
nally)
plus 1 fin (divided
by low midsec-
tion)
divided)
times modi-
fied)
# anal fins (external-
1
1 (divided by low
1
1(2)
1
1
ly)
midsection)
First fin to form
Vertical fins (pelvic
Pelvic (or anterior
?
Pelvic
Pelvic (except
Pelvic
rays
late)
dorsal ray)
Lota)
# pelvic rays
4
3-4 (larvae)
5-7 (adults)
5-7
2-11
3-4 (larvae)
5-8 (adults)
3-4 (larvae)
2-9 (adults)
Pelvic fin elongate?
No
Yes
No
Yes (most)
Yes
Yes
Pectoral ray forma-
7
Late
7
Midsequence
Late
Late
tion
to late
Body shape
Moderately elongate
Elongate
Moderately
elongate
Tapers to nar-
row peduncle
Elongate
Stocky
# vertebrae
64-69
43-59
58-62
41-63
59-84
44-55
Egg diameter (mm)
1.5-1.6
?
7
0.52-1.16
1.3-1.5
0.63-0.98
Chorion
Smooth
?
7
Smooth
Pitted/smooth
Smooth
# oil globules
1
?
7
1
1
Multiple to 1
Miscellaneous
Premaxillary spines
Small eye;
Barbel forms
Barbel forms
Barbels form on
(at least 1 form);
dense pig-
on lower
on lower jaw
lower jaw (and
barbel forms on
ment
jaw; few lu-
in juveniles
snout in
lower jaw
minescent
some); pterotic
spines in few
Gadidac
Merlucciidae
C« A. rt .-1 » ^
inenidae
Macroundae
Gadinae
Merlucciinae
Macruroninae
# dorsal fins (external-
3
2 (second divided 2
2
(1)2
ly)
by low midsec-
# anal fins (externally) 2
First fin to form rays
# pelvic rays
Pelvic fin elongate?
Pectoral ray formation
Body shape
# vertebrae
Egg diameter (mm)
Chorion
# oil globules
Miscellaneous
Caudal (pelvic
last)
6-7
No
Late
Moderately
elongate
39-64
1.0-1.9
Smooth
None
Barbel forms on
lower jaw dur-
ing/after juve-
nile stage
tion
1 (divided by 1
midsection)
Caudal (pelvic
7
Moderately
Late
Elongate
48-58
0.8-1.2
Smooth
1
1 (anterior rays elon-
1 (anterior rays elon-
1
gate in Macruron-
gate)
us)
Dorsal and anal (pel-
9
Pelvic (with dorsal and
vic late)
anal)
7-9
8
(0)5-17
No
No
Moderately to very
Late
Late (pedunculate)
Late
(pedunculate)
Attenuated (reduced
Attenuated (no caudal
Attenuated (no caudal
caudal fin)
fin)
fin)
77-78
7
80-116 +
0.99-1.16
7
1.0-2.0
Smooth
1
7
9
Hexagonal pattern
Luminescent organ
Some luminescent;
present; lacks caudal
barbel forms on low-
fin
er jaw, lacks caudal
fin
Fig. 138. (A) Muracnolepis sp., 32.5 mm SL, British Antarctic Survey, 53°48.7'S, 38°18.7'W. (B) Melanonus sp., 30.6 mm SL, MCZ 58619,
35°19'S, 07°30'E. (C) Svewvidovia. 13.0 mm NL, Fahay, 1983. (D) Svetovidovia, 44.4 mm SL, HML H6901, 38°49.5'N, 54"'18.0'W. (E)
"Svelovidovia." 44.1 mm SL, HML H9455, 43°21.94'N, 60°32.34'W.
FAHAY AND MARKLE: GADIFORMES
271
'^X/^^^^^//y^y^^^^^^
^$$i$s^i^^-
272
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 139. (A) Eretmophorus kleinenbergi. 105 mm, Mazzarelli. 1917. (B) Rhynchogadus hepaticus, 21.9 mm. Cipria, 1927. (C) Mora moro,
12 mm, De Gaetani, 1926. (D) Gadella maraldi. 18.8 mm, Sparta, 1928. (E) Physicutus nematopus, 9.2 mm, CALCOFI 5604, Sta. 103 G 40. (F)
Physiculus nematopus. 14.1 mm, ventral view, CALCOFI 5604, Sta. 103 G 40.
from the anus, small pedunculate pectoral fin, silvery eye and are unknown. Merlucciids have moderately pedunculate pec-
lack of caudal fin. torals; Merluccius approaches the gadines in pigmentation and
sequence of fin formation (caudal first), while macruronines
Merlucciidae. — Egs,s, larvae and juveniles of Mfr/wcaiw are well approach the macrourids in pectoral morphology and reduction
described (Table 73), while those of Lyconus and Macruronus of caudal fin.
Fig. 140. (A) Macrouridae, 1 1.2 mm TL, HML uncat., off Newfoundland. (B) Gadomus sp., 30+ mm TL, MCZ 58621, 25°48'N, 91°40'W.
(C) Corvphaenoides sp., 39 mm TL, MCZ 58622, 40°04'N, 68°07'W (pectoral fin damaged). (D) Coryphaenoides sp., 30+ mm TL, MCZ 58623,
34°27'N, 71°19'W.
FAHAY AND MARKLE: GADIFORMES
273
274
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 75. Meristic and Other Selected Characters tn Genera of Macrouridae. "Number of species" refers to minimum nominal species
followed by numbers surveyed for characters. Primary sources of data: Gunther, 1887; Gilbert, 1893; Goode and Bean, 1896; Gilbert, 1905;
Gilbert and Burke, 1912; Gilbert and Hubbs, 1916; Gilbert and Thompson, 1916; Koefoed, 1927; Hubbs, 1934; Parr, 1946; Koefoed, 1953;
Smith, 1961; Scott, 1962; Iwamoto, 1966; Makushok, 1966; Okamura, 1970b; Hart, 1973; Marshall, 1973; Marshall and Iwamoto, 1973; Iwamoto,
1974; Iwamoto and Stein, 1974; Hubbs and Iwamoto, 1977; Iwamoto, 1978; Merrett, 1978; Iwamoto, 1979; McCann and McKjiight, 1980;
Trunov, 1981; Merrett et al., 1983.
Nature of second
Longest rays
Number of
Retia
Branchio-
Precaudal
First
Pectoral
spinous ray of
(dorsal, anal,
Genus
species
Mirabilia
stegal rays
vertebrae
Pelvic rays
dorsal rays
rays
first dorsal fin
about equal)
Trachyrhynchinae
Trachyrhynch us
4/3
2
7
14
6-7
9-12
18-26
(ray)
D
Macrouroidinae
Macrouroides-
1/1
3
7
_
0
_
22-25
(ray)
Dor s
Squalogadus-
1/1
3
7
12-13
5
-
23-26
(ray)
Dor s
Bathygadinae
Gadomus
11/9
4
7
11-13
8-9
11-14
13-25
Smooth
D
Bathygadus
14/10
2
7
11-13
7-11
8-13
10-20
Smooth
D
Macrourinae
Cynomacrurus-
1/1
0
6
—
7-8
—
15-17
Smooth
A
Ondontomacrurus'^
1/1
2
6
—
7-8
8-10
8-11
Smooth
A
Lepidorhynchus
1/1
9
6
—
8-9
11-13
16-17
(19)
Smooth
—
Mahia
1/1
?
6
_
7
11
17-18
Smooth
A
Coelorhynchus
58/40
4
6
11-12
7
9-14
(13, 14)
15-21
(22)
Smooth
A
Hyomacrums
2/1
2
6
13
8
—
—
Serrate
—
Coryphaenoides
46/24
4
6
11-16
(6)
7-11
(12)
9-14
15-25
Serrate
A
Macrounis
3/2
4
6
16
8-9
10-13
17-21
Serrate
A
Nematonums
4/1
5
6
15
(7)
9-11
10-12
18-21
Serrate
A
Chalinura
8/6
6
6
12-13
8-14
9-12
18-22
Serrate
A
Lionurus
3/2
6
6
11-13
(15)
8-11
10-11
15-20
Serrate or
smooth
A
Mesobius'-^
2/2
2
7
11
6-9
10-12
12-16
Weakly
serrate
A
Echwomacrurus-
2/2
0
7
—
9-12
11-13
16-19
Serrate or
smooth
A
Hymenocephalus
17/3
2
7
10-11
7-15
10-13
(14)
11-18
(18-22)
Weakly
serrate
or smooth
A
Cetonurus
2/2
2
7
10
8-10
9-12
16-19
Serrate
A
Paracetonurus
4/3
2
7
11
(5)
6-9
9
8-11
17-21
Serrate
A
Kumba
1/1
2
7
-
11
21
Smooth
A
Parakumba
1/1
7
7
12
10
11
22
Weakly
serrate
A
Macrosmia
1/1
2
7
12
11-12
11-13
22
Weakly
serrate
A
Mataeocephalus
6/3
2
7
13-14
7-9
10-12
(19)
22-26
Serrate
A
Trachonurus
1/1
2
7
12-13
7
7-11
13-18
Smooth
A
Sphagemacrurus '
7/2
2
7
11-12
(8)
10-13
12-13
18-22
Serrate
A
Nezumia
44/32
2(4)
7
13-14
6-17
10-15
13-27
Serrate
A
Pseudonezumia
1/1
n
8
_
6
10
16
Serrate
A
Mataeocephalus'
4/3
2
7
14
8-10
11-16
16-22
Smooth
A
Venlrifossa
16/10
2
7
10-14
8-11
(13-15)
11-15
18-27
Serrate
A
' Juvenile phase known lo have prominently-spoued pigment pattern (but see note on Sphagemacrurus in Table
■ Includes bathypelagic species.
73).
FAHAY AND MARKLE: GADIFORMES
Table 75. Extended.
275
Mouth posnion
Number of hghl (terminal, sub-ter- Chm barbel
organs (bulbous minal. (present
or tubular) mfenor) or absent)
Position of anus and urogenital
opening relative lo anal fin
ongin and pelvic fin bases
A.
ong-
Plv.
bases
Anal fin ongin
antenor to
postenor end
of gut cavity?
Anteriormost
fin ongin
Distance between dorsal fins
<DI
base
= DI
base
>DI
base
Inf
Pr
No
D2or s
0
Inf
Ab
X
No
D2
Single
fin
0
Inf
Ab
X
No
D2
Single
fin
0
Term
Pr/Ab
X
No
D2
X
0
Term
Pr/Ab
X
No
D2
X
0
Term
Ab
X
IB
Term
Ab
X
—
A
X
IB
Term
Pr
X
-
-
IB
Sub-T
Pr
X
Yes
A
X
IB
Inf
Pr
X
or
X
Yes
A (usu.)
X
or X
or
X
or
(usu.)
IT
0
Inf
Pr
X
Yes
_
0
Sub-T,
Inf
Pr
X
Yes
A
X
or X
or
X
0
Inf
Pr
X
Yes/no
D2
X
0
Sub-T,
Inf
Pr
X
Yes/no
A
X
0
Sub-T
Pr
X
Yes/no
A
X
or
X
0
Inf
Pr
X
Yes/no
A
X
IB
Sub-T
Ab
X
Yes
A
X
0
Inf
Pr
X ? X
-
A
X
IT
Sub-T
Pr/Ab
X
No
A
X
Oor IB
Sub-T
Pr
(tiny)
X
Yes
A
X
0
Sub-T,
Inf
Pr
X
Yes
A(?)
IB
Sub-T
Pr
(tiny)
X
Yes
A
X
1B(?)
Sub-T
Pr
(tiny)
X
—
A
X
0
Sub-T
Pr
X
Yes
A
X
0
Inf
Pr
X or X
Yes
A
X
Oor IB
Sub-T
Pr
X
Yes
A
X
IB
Sub-T
Pr
X
Yes
A
X
IB
Sub-T
Pr
X
(X)
Yes
A
X
or
X
0
Inf
Pr
X
_
A
X
IB
Sub-T
Pr
X
Yes
A
X
IB
Sub-T,
Inf
Pr
X
(X)
Yes
A
X
or
X
276
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 141. (A) Macrouridae (Macrourinae). 60 mm XL, HML H6818, 39°52.5'N, 58°54.0'W. (B) Macroundae (Macrounnae), 1 5 mm TL, MCZ
58624, ag-SS'N, 79°54'W.
FAHAY AND MARKLE: GADIFORMES
277
~>«^
it
n
J pi
^'
-sjft'S"-'
^■^...JM^.X
.p^-
Fig. 142. (A) Steindachnena argentearlA mm TL, GCRL 01962. 28°45'N, 89°36'W. (B) Steindachneria argentea. 24 mm TL, GCRL 01962,
ventral view. (C) Mesobius berryi. 23.4 mm TL. Hubbs and Iwamoto, 1977.
Gadidae— The early life history stages of gadids are well known
(Table 73) and are reviewed and characterized in this volume
by Dunn and Matarese.
Selected Characters
•Eg?-?.— Eggs are undescribed for three gadiform families: Breg-
macerotidae, Melanonidae and Steindachneriidae. Efremenko
(1983b) recently described muraenolepidid eggs and Markle
(1982) summarized information for the remaining families and
noted that a relatively small egg (< 1 mm) with an oil globule
was a widespread and probably primitive character. The oil
globule has apparently been lost only in the gadines, a group
showing numerous derived stales, including relatively large eggs
(Markle, 1982).
Except in the gadid, Brosme, and macrourids, chorion or-
namentation appears to be restricted to ubiquitous pores seen
with scanning electron microscopy (Lannig and Hagstrom, 1975).
In B. brosme the chorion pores are many times larger than in
other gadiforms and give the egg a pitted appearance (Markle
and Frost, MS). In macrourids an elaborate "honey-comb" or-
naments the chorion. This ornamentation, like the pores, has
an unknown function. The uniqueness of the "honey-comb"
(Boehlert, this volume) and its presence in all known macrourine
eggs suggests an autapomorphy, at least for the subfamily. Ad-
ditional information on egg morphology of merlucciids, ma-
crourids, morids and Steindachneria could contribute to a dis-
cussion of the unsettled status and relationships of the latter.
Transient early life history characters— After hatching there are
at least six characters that can be considered ontogenetically
Fig. 143. (A) Gadiis morhua. 1 1.0 mm, Fahay. 1983. (B) Brosme bwsme. 14.0 mm, Fahay, 1983. (C) Urophycis chuss. 9.5 mm, Fahay, 1983.
(D) Merluccius productus. 10.1 mm, Ahlstrom and Counts, 1955.
FAHAY AND MARKLE: GADIFORMES
279
transient; larval pigmentation, lateral maxillary fangs, pterotic
spines, pedunculate pectoral fin bases, sequence of develop-
mental events and the presence of a pelagic juvenile stage. Many
of these characters are incompletely known for the order and
only tentative phylogenetic statements can be made.
Embryonic and larval pigmentation patterns are quite vari-
able. In gadoids there appears to be widespread occurrence of
postanal bands, usually one or two, and melanophores at the
notochord tip. Similar patterns occur in Merlnccius bilinearis
(Merlucciidae), Physiculus capensis (Moridae) and Coelorhyn-
chus sp. (Macrouridae). However, even within one subfamily
such as the gadines, there are genera without any banding (e.g.,
Melanogranuniis) as well as much variation in number of bands
(Pollachius. Gadus). Eye pigmentation at hatching varies de-
pending on development stage at hatching, for example, unpig-
mented in Pollachius and pigmented in Gadus. Embryonic and
larval pigmentation seems variable in the well studied gadids
as well as in other gadiforms so that an evaluation of phylo-
genetic significance seems premature at this time.
Lateral premaxillary spines are only known in Muraenolepis
(Fig. 1 38A) and larval pterotic spines are only known in Phycis,
Gaidropsarus (Demir, 1982; Markle, 1982) and Ciliata (Dunn
and Matarese, this volume). Both characters appear to be apo-
morphies, providing phylogenetic information at the generic
level at least. The western Atlantic Phycis. P. chesteri, lacks
larval pterotic spines and may, in fact, belong in Urophycis
(David Methven, pers. comm.).
The lack of developmental series outside the gadoids pre-
cludes discussion of many developmental sequence characters.
However, it does seem possible to make some tentative state-
ments about the first fin to form rays. On the basis of our
examination of a larval series provided by A. W. North of the
British Antarctic Survey, Muraenolepis does not form pelvic
rays first. This contrasts with most gadiforms where the pelvic
is the first fin to form (Table 74). Other exceptions seem to be
in gadines and merlucciids where the caudal or dorsal and anal
fins form before the pelvic. The latter condition may represent
a derived character state. However, the tail-less macrourids are
precluded from showing this character.
The pectoral fin base is strongly pedunculate (stylopterous)
during the larval period in macrourids and steindachneriids,
moderately pedunculate in morids and narrow-based (but less
pedunculate) in bregmacerotids, Meriuccius and gadids (Fig.
143A-D). Strong expression of this character is associated with
loss of the caudal fin (macrourids. steindachneriids) or delayed
caudal fin formation (morids) and may reflect a compensatory
response of larvae associated with larval locomotion.
In the life history of most gadiforms there is a benthic or
engybenthic adult phase. In all of these groups (muraenolepi-
dids, morids, most gadids, merlucciids, most macrourids) as
well as in pelagic gadiforms there is a prolonged pelagic juvenile
stage which, in some cases, includes symbiotic association with
jellyfish (Mansueti, 1963). In phycines, for example, this stage
is neustonic, includes a pigmentation pattern different from both
larval and benthic juveniles, and is characterized by a dense
concentration of melanophores on the dorsal surface. In morids,
some pelagic juveniles have been described as new genera, such
as Svetovidovia Cohen, 1973 (=Gargilius Koefoed, 1953).
We are not aware of any gadiform that can be shown not to
possess a pelagic juvenile. In fact, it appears that life-history
neoteny has occurred several times and adults have retained the
pelagic habitat (bregmacerotids, melanonids, the gadines Gad-
iculus and Micromesistius, and some macrourids). The pelagic
adult has clearly evolved independently more than once. Even
within a single family, Macrouridae, it has apparently happened
at least three times and Hubbs and Iwamoto (1977) have called
attention to this form of neoteny with the generic name, Me-
sobius ("middle life").
Pelvic fins.— The gadiform pelvic fin shows two major ontoge-
netic sequences. In the phycines, Urophycis and Phycis, larvae
initially form 3 or 4 rays and ontogenetically reduce or resorb
the innermost ray to produce the adult count of 2 or 3 (Markle,
1982). During the course of this study, we have also found
ontogenetic pelvic fin ray reduction in the morid Svetovidovia
vitellius. One transforming specimen, 55 mm SL (MCZ 59773),
has two large pelvic fin rays and 2 or 3 very minute remnants
of inner pelvic fin rays. Smaller specimens have as many as 1 1
rays (Table 72). Cohen (1979) has previously suggested that
Lotella ma.xillaris (10 pelvic fin rays) may be the young of
Laemonema (1-3 pelvic fin rays). Gadiforms may be pre-adapt-
ed for this type of metamorphosis since even in species with
numerous pelvic fin rays, such as the morid Physiculus, the
external fin ray nerves appear restricted to the outer two rays
(Freihofer, 1970; Fig. 12).
In the other, presumed ancestral, ontogenetic sequence, pelvic
fin rays increase in number. Variation is seen in this sequence
in the speed at which the adult complement is formed. The rays
form very quickly in Meriuccius, somewhat more slowly in Ele-
ginus. and over a protracted size range in Ga/^ropsarw,? (Markle,
1982; Dunn and Vinter, 1984, MS).
In many gadiforms, such as some macrourids, Meriuccius,
many gadids and morids, the pelvic fins also change allometri-
cally. In the Krohnius and several other types of macrourid
larvae as well as in morids, the pelvics are greatly expanded
over their relative size in any known adult. In some phycines
the pelvics are not necessarily relatively longer, but are wider
and fan-like as opposed to filamentous in adults. This allometry
favoring a relatively large, fan-like pelvic in the young would
seem to be a device to aid flotation. It is noteworthy, however,
that only bregmacerotids among the pelagic gadiforms have
retained enlarged pelvic fins as adults.
In addition to elongation, prominent pigmentation of pelvic
fins characterizes many genera, including most phycines. The
precise extent and location of pigment on the pelvic fin is often
an important identifying feature in these larvae. For example,
it is absent in Urophycis regia, restricted to the tips of the fins
in most other Urophycis and densely covers the fin membrane
in U. tenuis, Enchelyopus, Gaidropsarus and Raniceps.
Gadines, as previously mentioned, show a clear departure in
the sequence of fin formation. Instead of forming first, pelvic
fins form last. In Meriuccius, whose condition may be an evo-
lutionary precursor to the gadine condition, pelvic fins form
second in the sequence after the caudal (Table 74).
Pectoral fins. — \s is the case with most teleosts, pectoral fin rays
form late, although they may form before the late-forming cau-
dal in morids. As with the pelvic fin, the pectoral fin is often
elongate and/or fan-shaped in some morids and macrourids (i.e.,
Gadella and Hymenocephalus) but this fin is not prominently
pigmented in any member of the order except some species of
Meriuccius.
Dorsal and anal fins. — \n the development of all gadiforms de-
scribed, vertical fins form in their adult positions and there is
no evidence of fin base migration. The dorsal fin origin in gad-
280
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 76. Distribution of Caudal Rays on Supporting Bones of the Calidal Fin in Selected Species of Gadiformes. "Inferior Hypurals"
do not include the parhypural in this listing. See Table 72 for primary sources of data.
Epu-
rals
Hypu-
rals
XY
bones
Number of
vertebrae
associated
with
caudal Iin
Dorsal
Number of rays
Superi-
or Infenor
hypu- hypu-
ral(s) ral(s)
Ventral
Taxa
(Un-
branched)
(Branched)
(Branch-
ed)
(Un-
branched)
caudal
rays
Muraenolepididae
Muraenolepis sp.
2
2
See text
See text
2-3
4
2
1-2
8-10
Bregmacerotidae
Bregmaceros
balhymaster
2
2
Present
5-6
(10-12)
8
2
(10-12)
34-36
(?)
Melanonidae
Meianonus zugmayeri
2
(4) 5 (6)
Absent
13
23-25
6
3
22-25
55-60
Fam. Incertae sedis
Euclichlhys sp.
2
2
X present
Y absent
10
17
4
3
17
41
Moridae
Antimora rostrata
Brosmiculus sp.
Eretmophorus
kleinenbergi
Gadella sp.
Gadella maraldi
Laemonema barbatula
Laemonema longipes
Lepidwn eques
Lepidion lepidion
Mora mora
Lotella fernandeziana
Physiculus neniatopus
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
5(6)
5
5
5
5
5
5
5
5
5
5
5
5
5(6)
5
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
Present
8
9
7
7
4-5
5-6
7
5
7(?)
(5-6)
(6-7)
(6-8)
(4-6)
(6-8)
14
15
16
8-9
11-13
15
7
(7)
8-9
(2-3)
(6-7)
(3-4)
(3-5)
5
6
5
6
6
24-
6
5-6
13-
5-6
14-
6
6
6
4
4
4
3-4
3
(5-6)
12
13
13
12-15
15-17
(6-7)
(8-9)
(6-7)
(5-7)
(7-8)
(30)
35
38
26
38
20-24
27-31
22-25
34-36
22
3-4
4
(5-6)
(0-1)
34-38
22-26
Physiculus
rastrelliger
Salilota australis
Triplerophycis
gilchrisli
Svetovidovia
4
(1-3)
27-32
4
4
3
15
5
(12)
16-18
40
22
(28)
30-33
Lotinae
Brosme brosme
2
2
2
2
2
2
Absent
Absent
Absent
(variable)
14
13
19-20
4
3
22-24
45-48
Lota lota
Molva molva
(4-5)
20-22
22-23
(14-15)
4 2
5 3
(12-13)
21-25
25
(4-6)
42-50
52-53
Phycinae
Enchelyopus cimbrius
2
2
Present
8-9
(5)
12-15
(9)
(4)
5
2
(6)
13-15
(5)
31-35
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Present
Present
Present
Present
Present
Present
Present
8-9
7-8
6
7
8
(6-8)
(3)
(7-8)
(9)
-24
2
*
3
(3-4)
18-21
14-15
14-17
(6-8)
(4)
(7-8)
(9)
Gaidropsarus ensis
Gaidropsarus
mediterraneus
Phycis chesteri
Phycis blennioides
Urophycis regia
Urophycis chuss
Urophycis tenuis
17-20
13-15
12-14
11-12
12-13
13-15
(5-6)
(4-5)
(5)
5
(4)
5
6
6
6
(5)
6
41-46
31-36
32-37
28
30-32
29-34
33-39
3
3
3
(2)
(3)
13-14
10-11
13-14
Gadinae
Arctogadus borisovi
Boreogadus saida
Eleginus gracilis
Gadiculus argenteus
2
2
2
2
2
2
Absent
Absent
Absent
15
13-14
15-16
(15)
(7-8)
21-26
21-25
22-25
(8)
4
4
5
20-
1
2
(3)
2
(6)
22-25
21-25
23-26
(16)
(7-8)
47-55
46-54
50-56
35-36
(?)
FAHAY AND MARKLE: GADIFORMES
Table 76. Continued.
281
Epu-
rals
Hypu-
rals
XY
bones
Number of
vertebrae
associated
with
caudal tin
Dorsal
Num
Superi-
or
hypu-
ral(s)
ber of rays
inferior
hypu-
raKsl
Ventral
To
cau
ra
tal
Taxa
(Un-
branched)
(Branched)
(Branch-
ed)
(Un.
branched)
dal
Gadus morhua
Gadus ogac
Gadus macrocephalus
Melanograminus
aegtefimis
Microgadus tomcod
Microgadus proximus
2
2
2
2
2
2
2
2
2
2
2
2
Absent
Absent
Absent
Absent
Absent
Absent
13-14
14
11-13
14
13-14
14-15
(12-13)
(15-16)
(9)
22
21
25-
21
■yi.
-26
-23
-27
-22
-75
(11)
(11)
(12)
(11-13)
4
4
4
5
4
(4)
2
2
2
2
1
(9)
(9)
(10)
(9-11)
(13-14)
(8-9)
23-
21-
26-
21-
-27
-24
-27
-24
-''4
(12-13)
(15)
(10)
(11-12)
(16-17)
(11-13)
49-
46-
51-
57-
46-
49-
-57
-51
-53
-60
-50
-56
Micromesistius
poulassou
Poltachius virens
Theragra
chalcogramma
2
2
2
2
2
2
Absent
Absent
Absent
15
15-16
12-13
(15-
(12-
-16)
-14)
21-
30-
-22
-32
(15-16)
(9-10)
5
5
5
4
2
2
2
23-
31-
-24
-33
50
66-70
Merluciidae
Merlucaus albidus
Merlucinis bilinearis
2
2
2
2
Present
Present
10
9
(10)
(8)
16
13-15
(6)
(5-6)
6
6
3
(2)
3
2-3
2
(4)
(5-6)
18
15-17
(10)
(7-8)
40
34-37
Merluccius productus
Macruronus
novaezelandiae
2
2
2
Present
See text
9-10
2
(10-
-11)
4?
(6-7)
6
3
(4-6)
2-
-3?
(10-11)
42-
10
-43
-12?
• Included in ventral count.
iforms vaiies from occipital to slightly behind the pectoral fin-
tips and is long-based, reaching almost to the caudal fin (and
confluent with the caudal in Muraenolepis and Macruronus).
The anal fin originates close behind the anus (except in some
macrourids and steindachneriids) and also extends posteriorly
to near the caudal fin (confluent with caudal in Muraenolepis
and Macruronus). The caudal peduncle is, as a result, very short,
being longest in Merluccius. Variations are fijund in morids
(where the dorsal fin may extend farther posteriorly than the
anal, accommodating an asymmetrical caudal fin), and in Stetn-
dachneria and the Macrouridae (where the lack of a caudal fin
results in dorsal and anal rays tapering until they meet at the
tip of the tail).
Gadiform dorsal and anal fin rays usually form after pelvic
fin rays begin (with the exception of the anterior dorsal ray in
Bregmaceros. which may form very early). Usually, the vertical
fins ossify together, so that caudal, dorsal and anal fin rays
appear at about the same time. In some genera with one long
dorsal fin (or a short first dorsal fin preceding a longer second
dorsal) fin rays in the longer fin form from two or more centers
of ossification, for example, Molva (Schmidt, 1906b) and Ater-
luccius (Ahlstrom and Counts, 1955). This may indicate either
preadaptation for the multiple dorsal/anal condition or second-
ary loss of multiple dorsals/anals.
In cases where a long second dorsal fin is preceded by a rel-
atively short first dorsal fin, development of the first is usually
delayed and is often the site of the last fin ray formation, as
in Gaidropsarus {Gcmn. 1982) and tVop/jic/.v (Hildebrand and
"^able, 1938). In gadines, which have three dorsal fins, the first,
again, forms last. In Macrouridae and Merluccius. however, first
dorsal fin rays appear to form much sooner than the second
(Merrett, 1978; Ahlstrom and Counts, 1955). In the morids
Physiculus and Svetovidovia the first and second dorsal fins
develop together and divide late in development into apparent
first and second fins. Thus in the gadiforms, ossification of ver-
tical fin rays is not uniformly anterior to posterior, or vice-versa,
or from the middle toward both ends, but instead is variable.
The relationship of dorsal and anal fin rays to centra is an
important characteristic in gadiforms (Rosen and Patterson,
1969; Marshall and Cohen, 1973). We believe the primitive
state, as exemplified by Muraenolepis, involved about three rays
per centrum. The major evolutionary trend, as identified by
Rosen and Patterson ( 1 969), is for this ratio to change anteriorly
where it is replaced by an approximate 2:1 relationship in most
families of gadiforms. In phycines, two derived character states
are found: a 7:1 relationship in rocklings (Cohen and Russo,
1979; Markle, 1982: fig. 5B) and a 2:1 relationship in hakes
(Markle, 1982: fig. 5C).
Caudal fin. — Y>e^X>^\t its absence in over a third of all gadiforms,
the distinctive caudal skeleton has received an inordinate amount
of attention. In the present context it is doubly important, for
its presence offers identification as well as phylogenetic infor-
mation (Markle, 1982; Dunn and Vinter, 1984, MS). Tailed
gadiform larvae typically have a symmetrical fin where some
"secondary" rays form before some "primary." As used here,
"primary" refers to those rays articulating with the superior
hypural (hypurals three through five) while rays attached to
inferior hypurals (hypurals one and two), parahypurals, epurals,
accessory bones (the X and Y bones of some authors), or to
elongate neural and haemal spines are referred to as "second-
ary." Secondary rays also include those originating between
neural or haemal spines.
We have summarized the distribution of gadiform caudal fin
rays in Table 76. Various authors lump secondary rays together
or express them as branched or unbranched. We have included
both methods, thus the sum of counts from different sources do
not always correspond. Several things are apparent from this
table. One involves the utility of using the distribution of caudal
rays both in intra- and intergeneric comparisons (Markle, 1982;
282
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
CFR
A Muraenolepis
Fig. 144. Hypothesized acquisition of gadoid caudal structure from
condition in Muraenolepis (see text). X and Y bones shaded. Abbre-
viations: CFR: caudal fin rays; EP: epurals; HS: haemal spines; IH:
inferior hypural; NS; neural spine; PH: parhypural; PU,: first preural
centrum; R; radials; SH; superior hypural; U,: first ural centrum; U,;
second ural centrum.
Dunn and Vinter, 1984, MS). Primary and total caudal fin ray
counts also exhibit some difference in symmetry and patterns
of evolutionary change. Morids are the only group of tailed
gadiforms that show noticeable asymmetry in superior versus
inferior secondary caudal rays (Table 76). Morids and some
phycines have relatively low total caudal fin ray counts (20-38)
and numerous groups have some genera with six primary caudal
fin rays. Markle (1982) interpreted both of these as primitive
states.
The caudal oi Muraenolepis differs from most other gadiforms'
in its complete continuity with both dorsal and anal fin rays
(Fig. 144A). It is virtually identical to that of the ophidiiform,
Brotula (Monod, 1968), differing only in number of rays sup-
ported by the parhypural (one vs. two). The typical gadoid
caudal fin skeleton is easily derived from the condition in Mu-
raenolepis, which we identify as the primitive state. The primary
requirement is the acquisition of X and Y bones and modified
spines of the first preural centrum, both sets of which must have
cartilaginous articulating surfaces entering into support of cau-
dal rays. X and Y bones are present in Muraenolepis as the
penultimate radials of the dorsal and anal fins. If the last radials
fuse with the spines of the first preural centrum, both sets of
preural caudal bones (with cartilaginous articulating surfaces)
are acquired.
A second requirement is an interspace (lacking rays and ra-
dials) between the caudal fin and the dorsal and anal fins. This
condition could have been satisfied in one of two ways. Rays
(and their supporting radials) anterior to the X and Y bones
might have been lost, and subsequent changes in caudal ray
numbers would then involve the addition of secondary rays
lacking radials. A less parsimonious scenario involves the loss
of radials (only) anterior to the X and Y bones which leaves a
continuous dorsal-caudal-anal fin including some anterior un-
supported rays. In this case, further variation in numbers of
secondary caudal rays would involve both increases and de-
creases. The hypothesized ancestral gadoid condition is shown
in Fig. 144B. Presumably, this ancestor would have had 16
caudal fin rays (one each on X and Y bones, first preural neural
and haemal spines, each epural and parhypural, six on the
superior hypural and three on the inferior hypural). This total
is close to the lowest known (and presumably most primitive)
counts in certain morids (Table 76) and corroborates the sug-
gestion that higher counts in Melanonidae, Gadidae and Mer-
luccius are derived states brought about through the acquisition
of additional rays lacking radial support (Fig. 144C). In Brosme
this acquisition has resulted in a secondary elimination of the
caudal peduncle and an almost continuous dorsal-caudal-anal
fin (Markle, 1 982: fig. 7C). The acquisition of rays has apparently
occurred asymmetrically in some morids, where ventral sec-
ondary rays outnumber dorsal.
Olfactory lobes.— J\\e position of olfactory lobes relative to na-
sal organs and the forebrain has been used as a systematic char-
acter in gadiforms by Svetovidov (1948, 1969) and Marshall
( 1 965). This character develops during ontogeny since the bulbs
are close to the forebrain in young of all gadiforms (Rass in
Svetovidov, 1948; Marshall, 1965). It reaches the most derived
state adjacent to olfactory capsules in "nearly all of the Gadi-
dae," "most species of Macrouridae," Muraenolepis (Marshall.
1966b) and Merluccius (Inada, 1981b). Olfactory lobes are be-
tween the forebrain and olfactory capsules in Bregmaceros and
next to the forebrain in other merlucciids and Steindachneria
(Marshall, 1966b).
We are not certain how to interpret the available information
' The caudal fin of Macruromis novaezelandiae. though much reduced
in over-all size, is similar to that of Muraenolepis \n its confluence with
dorsal and anal fins.
FAHAY AND MARKLE: GADIFORMES
283
on this structure since an undescribed ontogenetic sequence is
involved. This character is an important part of our current
concept of Merlucciidae, thus descriptions of its ontogeny could
contribute to a better understanding of this family's interrela-
tionships.
Genilal papilla.— A. genital papilla develops precociously in most
gadiforms. It is most pronounced in morids. macrourids and
Steindachneria (see figures), but we also could find it in gadids
and Merluccius.
Mental barbels. — Mental barbels usually develop late in the lar-
val or early in the pelagic juvenile period. They are found in
most gadids (being lost in some of the secondarily pelagic forms
such as Pollachhis pollachiiis and Micromesistiiis). most ma-
crourids, muraenolepidids and morids. Additional fleshy, snout
barbels are found in phycine rocklings. The propensity to de-
velop snout and mental barbels seems widespread in gadiforms
and can also be found in some ophidiiform fishes; it appears to
have a strong ecological component and we are unable to attach
phylogenetic significance to its presence or absence.
Gadiform Phylogeny
A framework of interrelationships of the eight gadiform fam-
ilies has developed from among others, Marshall ( 1 966b), Gos-
line (1968), Rosen and Patterson (1969), Okamura (1970b) and
Marshall and Cohen (1973). A concensus on minor as well as
some major points does not exist, and we therefore follow a
modification of Rosen and Patterson (1969) and Cohen (this
volume). In this framework muraenolepidids are the most prim-
itive group, showing no obvious relationships, and are the pre-
sumed sister group to all other gadiforms. Based on fossil
evidence (Danil'chenko, I960), bregmacerotids are thought to
be related to a group composed of morids and melanonids.
These three families are the sister group of macrourids and
together form a principal gadiform lineage. Steindachneria and
merlucciids are sister groups and with gadids form the other
principal gadiform lineage.
On the basis of available data, we can identify the following
early developmental characters, their denved states, and known
distribution in the order. In many cases the "holes" in our data
severely reduce the weight of our arguments. (1) Oil globule in
egg— lost — gadines; (2) Chorion ornamentation — honey-
combed—macrourines; (3) Lateral premaxillary spines— pres-
ent—muraenolepidids; (4) Pterotic spines— present— some phy-
cines; (5a) Sequence of fin formation— caudal first— gadines and
Merluccius; (5b) Sequence of fin formation — pel vies last— gad-
ines; (6) Pelvic fin ontogeny— reduction in ray number— phycine
hakes and morids; and (7) Larval pectoral fin— pedunculate—
macrourids and Steindachneria. To this list we can add onto-
genetically persistent characters taken in part from Rosen and
Patterson ( 1 969), Marshall and Cohen (1973) and Markle (1982).
(8) X and Y bones— loss in forms with tails— melanonids, gad-
ines and lotines; (9) Total caudal fin rays— over 50 — melan-
onids, gadines and lotines; (10a) Anterior dorsal fin rays to
centra ratio — 7:1 —phycine rocklings; (10b) Anterior dorsal fin
rays to centra ratio— ca. 1:1— gadines, morids?, macrourids and
merlucciids: (1 1) Precaudal vertebrae— counts greater than 20—
gadines, lotines, merlucciids and muraenolepidids; (12) Hypur-
als— fusion into two plates— muraenolepidids, bregmacerotids,
gadids and meriucciids; (13) Otophysic connection — present—
morids; and (14) Fin diflferentiation— three dorsals and two
anals— gadines, some morids (Merluccius and bregmacerotids
to a lesser degree).
These characters generally do not support the above hypoth-
eses of relationships. Notable discrepancies and areas for ad-
ditional investigation are: ( I ) whether gadids are monophyletic,
specifically whether phycines belong in and Merluccius belongs
out; (2) relationship, if any, of melanonids to gadines; and (3)
relationships of Steindachneria.
(M.P.F.) National Marine Fisheries Service, Northeast
Fisheries Center, Sandy Hook Laboratory, High-
lands, New Jersey 07732; (D.F.M.) Huntsman Marine
Laboratory, Brandy Cove, St. Andrews, New
Brunswick EGG 2X0 Canada.
Gadidae: Development and Relationships
J. R. Dunn and A. C. Matarese
LARVAE of the fishes of the family Gadidae have received
a great deal of study through the years and because of the
economic value of the family, the larvae are taxonomically as
well known as those of most families of teleosts. Svetovidov's
(1948) classic work on the systematics of adult gadid fishes is
the benchmark of knowledge of the family. He considered 22
genera (including Merluccius). examined osteological characters
of representatives of all genera, and based his classification
scheme mainly on the structure and number of median fins (see
also Svetovidov, 1956). Subsequent workers (Mujib, 1967, 1969;
Marshall and Cohen, 1973) have extended our understanding
of the relationships of certain members of the family, but a
comprehensive study of Gadidae, including early life history
stages, has not yet been accomplished. Recently Markle (1982)
examined larval and adult representatives of all gadoid families
which led him to recognize three gadid subfamilies: Phycinae,
Lotinae, and Gadinae.
Our purpose here is to summarize available knowledge of the
taxonomy of eggs and larvae of the family Gadidae. We include
observations on eggs, larval morphology and pigment patterns,
and developmental osteology. Included are illustrations of lar-
vae of representatives of all currently recognized gadid genera
284
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 77. Summary of Egg Characters in Genera of the Family Gadidae. All eggs are spherical in shape with a homogeneous yolk.
Oil globules
Pelagic/
Size
Taxon
demersal
(mm)
Number
Size (mm)
Pigment
Lotinae
Brosme
P
1.29-1.51
1
0.23-0.30
Molva
P
0.97-1.13
1
0.28-0.31
yes
Lota
D
1.00-1.90
1
yes
Phycinae
Enchelyopus
P
0.66-0.98
multiple to !
0.13-0.20
yes
Gaidropsarus
P
0.70-0.85
multiple to i
0.14-0.16
yes
Phyas'
P
0.76-0.79
1
Urophycis
P
0.63-0.97
multiple to '.
0.17-0.20
yes
Citiata
P
0.67-0.98
multiple to 1
0.11-0.18
yes
Rankeps
P
0.75-0.91
1
0.14-0.19
yes
Gadinae
Trisopterus
P
0.90-1.22
0
Merlangius
P
0.97-1.32
0
Pollachius
P
1.00-1.22
0
Melanogrammus
P
1.10-1.72
0
Gadus
P. D
0.92-1.90
0
Gadiculus
0.91-0.97-
Microgadus
D
1.39-1.70
probably 0
Eleginus
D
1.00-1.70
0
Boreogadus
P
1.53-1.90
0
Arctogadus
D
Theragra
P
1.19-1.81
0
Micromesistius
P
1.04-1.27
0
Russell (1976), Schmidt (1905b)
Schmidt (1905b)
Brederand Rosen (1966), Jude (1982b), Morrow
(1980), Schmidt (1907a). Snyder (1979)
Fahay (1983), Hardy (1978a). Russell (1976)
Dekhnik (1973), Russell (1976)
Wenner(1978)
Fahay (1983), Hardy (1978a)
Russell (1976)
Kennedy and Fitzmaurice (1969), Russell (1976)
Russell (1976)
Dehknik (1973), Russell (1976)
Fahay (1983), Fridgeirsson (1978), Russell (1976)
Hardy (1978a), Russell (1976)
Mukhacheva and Zviagina (1960), Russell (1976)
Russell (1976)
Hardy (1978a), this study
Breder and Rosen (1966), Kozlov (1952), Mukhacheva
(1957), Mukhacheva and Zviagina (1960)
Pertseva (1936). Rass (1968), Russell (1976)
Zviagina (1961)
Gorbunova (1954), Yusa (1954)
Lisovenkoet al. (1982), Russell (1976), Seaton
and Bailey (1971)
' Applies lo P. cfieslen only.
' Ovarian eggs only.
(except Arctogadus). Finally, we attempt to evaluate the rela-
tionships of the subfamilies of gadid fishes based on early life
history and adult characters.
Methods
We have examined developmental series of varying com-
pleteness of representatives of all gadid genera except Phycis
and Arctogadus. of which only juvenile specimens were avail-
able to us. Measurements were taken on these series and smaller
series were differentally stained (Dingerkus and Uhler, 1977)
for study of developmental osteology.
Russell (1976) described eggs and pigment patterns in gadid
larvae. Matarese et al. (1981) modified the terminology used by
Russell in describing postanal pigment and we use their ter-
minology here.
Our discussion of osteology presented here is limited pri-
marily to features of the pectoral and pelvic girdle, vertebral
column, and median and paired fins. Svetovidov (1948) and
Mujib (1967, 1969) have described cranial osteology. Matarese
et al. (1981) and Markle (1982) have discussed the significance
of median fins in gadoid fishes.
Characteristics of Adult and Early
Life History Stages
Family Gadidae.— Gadid fishes possess four to six pectoral ra-
dials; the posttemporal is attached to the skull in adults and
possesses a ventral branch of varying length. An anterior process
of varying length is present on the coracoid, but a posterior
process is lacking. The postcleithrum is variously curved and,
in some genera, possesses an expanded distal head. The pelvic
basipterygia have a posterio-lateral process of varying length.
The first neural spine is attached to the supraoccipital crest in
adult fishes; and subsequent anterior neural spines (on vertebrae
2-10) vary in length and are oriented vertically or posteriorly.
One or two predorsal bones are present in some genera but are
absent in most. One to three dorsal and one or two anal fins are
present. When two dorsal fins are present, the first may be
separate from the second (intemeural bones absent), or contin-
uous (intemeural bones present); when three dorsal fins are pres-
ent the second is always internally continuous with the third;
when two anal fins are present, they too are internally contin-
uous. The distance (number of intemeural bones) between mul-
tiple dorsal and anal fins varies among genera.
The caudal fin has three hypural bones (Matarese et al., 1981)
including the parhypural (Markle, 1982), and four to six "pri-
mary" caudal fin rays (those articulating with the superior hy-
pural bone [Markle, 1982]). Accessory (.x and y) bones are pres-
ent or absent, two epural bones are present, uroneural bones are
absent, two ural centra are present, and the neural and haemal
spines on preural centra one are broadly spatulate in most gen-
era. Matarese et al. ( 1 98 1 ) did not detect fusion of hypural bones
during ontogeny of Microgadus proximus, but they hypothe-
sized that hypural 2 represented a fusion of hypurals 2 and 3
and that hypural 3 represented a fusion of hypurals 4-6. because
of the presence of three inferior and three superior hypural
elements in Moridae (Fitch and Barker, 1972), which is generally
considered a more primitive family than Gadidae.
Subfamily Lotinae (Tables 77-82. Figs. 145-146.). — Membtrs
of the Lotinae are elongate gadid fishes including Brosme, Mol-
DUNN AND MATARESE: GADIDAE
285
Fig. 145. (A) Preflexion larva of Bros me brosme. 5.9 mm SL (Huntsman Mar. Lab., H-16260. stored at NWAFC); (B) Flexion larva of Molva
moha. 8.2 mm SL (Inst. Sci. Tech. Peches Marit., Nantes, stored at NWAFC); and (C) Preflexion larva of Lota lota. 3.7 mm SL (Group Interuniv.
Res. Oceanogr., Quebec, stored at NWAFC).
va, and Lota (MarkJe, 1982). Brosme is monotypic and occurs
on both sides of the North Atlantic Ocean. Molva. with three
nominal species, occurs in the east and west North Atlantic
Ocean (Svetovidov, 1 948; Leim and Scott, 1 966). Lola is mono-
typic and two subspecies occur in fresh and brackish waters of
Europe, northern Asia, and North America (Pivnicka, 1970).
The characteristics of the subfamily, based on Markle (1982)
and this study, are egg diameter relatively large (0.97-1 .90 mm);
oil globule present (0.2-0.3 mm diameter); vertebrae numerous
(62-66 total, 20-26 precaudal in specimens examined); pterotic
spines absent; pelvic ray formation prior to notochord flexion
but acquisition of adult complement delayed; x and v bones
usually absent; 4-5 primary caudal fin rays; 45-54 total caudal
fin rays; and numerous total dorsal and anal fin rays (77-108D
and 59-75A).
Eggs and larvae of lotines are reasonably well known (Tables
77-79). Brosme and Molva shed planktonic eggs whereas Lola
deposits nonadhesive, demersal eggs, all with a single oil globule.
The chorion of eggs of Brosme has deep pits visible by scanning
electron microscopy (Markle, pers. comm.').
Lotine larvae hatch at moderate sizes (3-4 mm), yolk is ab-
sorbed at around 5 mm, and notochord flexion is delayed (9-
25 mm). Size at transformation is large and the duration of the
pelagic stage is extensive (Table 78). The larvae tend to be
slender to moderately slender and taper toward the tail. Pelvic
fins are precocious in Brosme and Molva. but not Lola.
Head pigment in larvae is generally limited to the mouth and
dorsal area of the head. Gut pigment is sparse, initially located
only on the dorsal surface. Brosme and Molva have pelvic fins
' D. F. Markle, Huntsman Marine Laboratory, St. Andrews, New-
Brunswick, pers. comm., 25 February 1983.
which are pigmented distally. Postanal pigment patterns are
similar in Brosme and Molva (Table 79). Brosme larvae have
two postanal bars and distinctive pigment above and below the
urostyle (Fig. 1 45 A). Although Molva does not have a bar pat-
tern initially, the dorsal and ventral pigment eventually coalesce
into two postanal pigment bands, the characteristics of which
are of taxonomic value in differentiating species in the genus
(Fig. 145B). Preflexion larvae (3-7 mm) of L. lota (lacusiris?)
in North American waters were reported by Fish (1932) to lack
postanal pigment. Snyder (1979), however, reported finding dor-
sal and ventral postanal pigment in preflexion larvae identified
as L. lota, as we did in those we examined from James Bay,
Canada (Fig. 145C).
Brosme has single dorsal and anal fins with a slight separation
between the anal and caudal fins (Markle, 1982). The neural
spine on preural centrum one (PU,) is distally flattened, the
haemal spine on this centrum is distally rounded (Table 82),
and \/y bones are absent (Fig. 146). Molva possesses two dorsal
fins with only a slight internal separation. The haemal spine on
PU, is distally rounded and x/v bones are present or absent
(usually absent). Lola also possesses two dorsal fins, with only
slight internal separation, and a single anal fin. Both the neural
and haemal spines on PU, are distally flattened and the species
usually lacks x/v bones, but a reduced x and/or y bone is some-
times present (Markle, 1982).
Subfamily Phycinae (Tables 77-82, Figs. 147-148). -T^xt
subfamily Phycinae was resurrected by Markle (1982) who ex-
amined seven species of Northwest Atlantic gadids belonging
to four genera: Enchelyopus, Gaidropsarus, Phycis. and (Jro-
phycis. We include also Ciliata and, arbitrarily, Raniceps as
phycines. Enchelyopus and Raniceps are each monotypic; the
former is found on both sides of the North Atlantic Ocean, the
286
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 78.
Summary of Morphological Characters of Larvae of the Family Gadidae. Proportions are expressed as percentages of standard
length, when possible.
Standard length (mm
Morphomelncs (% SL)
Eye diameter
Head length
Body depth'
Post-
Pre-
Pre-
Posl-
Pre-
Post-
Pre-
Taxon
Halching
Rexion
flexion
juveniie
Juvenile
flexion
flexion
flexion
flexion
flexion
Lotinae
Brosme
4
14-25
25-40
40-60
>60
21-
16
Molva
3
9-14
14-20
20-80
>80
9
7
30
24
21
Lota
3
14-19
19-30
8
6
21-25
Phycinae
Enchelyopus
2
5-7
7-18
18-45
>45
10-15
4-5
Gaidropsarus
2
5-7
7-12
12-48
>48
11-13
9-13
20-30
20
30
Phycis
7
5-8
8-12
12-30
>30
9
23
Urophycis
2
4-5
5-
-40
>40
Ciliata
2
5-8
13
9
28-32
27
25-30
Raniceps
3
7-12
11
27
32
Gadinae
Trisopterus
3
7-11
>45
8
11
26
30
22
Merlangtus
3
9-13
13-23
9
9
28
27
20
Pollachius
3-4
12-16
>50
9
21
26
16
Melanogrammus
3-4
10-16
16-22
>90
10
Gadus
3-4
10-17
17-25
25-35
>35
5-9
7-10
11-27
22-26
10-18
Gadiculus
?
7-13
13-30
30-40
>40
10
31
Microgadus
3
8-15
14-28
28-46
>46
8
8
22
32
19
Eleginus
4
11-17
17-24
24-27
8
9
20
26
15
Boreogadus
6
11-17
17-30
30-45
>45
8
18
14
Arclogadus
?
28-
-31
3
22-25
16-19
Theragra
3-4
10-17
17-25
25-40
>40
7-
-8
15
33
Micromesislius
2
8-13
>32
9
9
29
32
26
' At pectoral fin base when possible.
^ Data between columns indicate no data available for ditferences in pretlexion and flexion larvae.
' Data are % HL.
latter in the eastern North Atlantic, as are the two nominal
species of Ciliata (Cohen and Russo, 1979). Gaidropsarus has
about 14 nominal species, 1 1 in the North Atlantic Ocean and
single species off South Africa, New Zealand, and Japan. Phycis
has three nominal species occurring in both sides of the North
Atlantic Ocean. About seven species of Urophycis are presently
recognized in the western Atlantic Ocean from Canada to South
America (Svetovidov, 1948).
Characteristics of phycines according to Markle (1982) and
this study include: egg diameter small (0.63-0.98 mm): multiple
oil globules that eventually coalesce into a single moderately
sized globule (0.1 1-0.22 mm diameter); vertebrae moderately
numerous (45-55 total, 14-17 precaudal); pterotic spines pres-
ent in some larvae and juveniles of Gaidropsarus, Ciliata. and
Phycis; initial pelvic fin ray formation prior to flexion but ac-
quisition of adult complement delayed; x and y bones present
(sometimes absent in Raniceps); 5-6 primary caudal fin rays
and 29-38 total caudal fin rays; and moderate numbers of total
dorsal and total anal fin elements [D, 49-73, not including spe-
cialized rays (e.g., Enchelyopus), and A, 40-57].
Taxonomic problems are prevalent in this group. Specific
identification of smaller larvae is not presently possible for cer-
tain species of Gaidropsarus. Phycis. and Urophycis (Russell,
1976; Markle, 1982). Adultsof some species are easily confused
(Musick, 1973; Svetovidov. 1982).
Eggs of Enchelyopus and Ciliata. some Urophycis. as well as
those oi Gaidropsarus that are known, have multiple oil globules
in the earliest stages, which coalesce into a single oil globule;
melanistic pigment is present on both the embryo and oil globule
(Table 77). In Phycis. only ovarian eggs of P. chesteri have been
described (Wenner, 1978). Multiple oil globules have not been
observed in U. tenuis (Markle-) nor reported in eggs oi Raniceps
raninus.
Phycine larvae hatch at small sizes (1.5-3.0 mm), yolk is
absorbed quickly, notochord flexion occurs at small sizes (about
5-12 mm), size at transformation is variable, and a prejuvenile
stage is present in most genera (Table 78). Phycine preflexion
larvae tend to be deeper bodied (at the pectoral fin base) than
lotines or gadines, but with development they become mor-
phologically diverse. Pelvic fins are precocious and pigmented,
although the extent and duration of pigmentation varies among
genera. The entire fin is pigmented in Gaidropsarus. whereas
only the tip is pigmented in some species of Urophycis. Raniceps
is morphologically the most divergent phycine. By 5 mm, the
preanal portion of the body is usually high in relation to the
postanal region and at a length of about 7.5 mm the larvae are
"tadpole shaped."
Although pigmentation is highly variable in phycines, most
genera possess head pigment on the dorsal part (sometimes
extending to the nape), and on the snout and mouth. In addition,
some genera may have pigment near the eye and on the opercular
area. Gut pigment is initially located along the dorsal surface,
with some genera (e.g., Enchelyopus. Phycis) developing more
pigment over the lateral surface. Postanal pigment is variable.
' D. F. Markle, pers. comm., 5 July 1983.
DUNN AND MATARESE: GADIDAE
287
Table 78. Extended.
Morphomelncs
(%SL)
Body depth
Preanal length
Approximate
Posi-
Pre-
Posl-
Lateral
Precocious
pelvic fin
flexion
flexion
flexion
vent
pelvic fin
fonnation
35
48
yes
yes
6
15
43
48
yes
5
47
yes
yes
no
yes
11
3
20
60
37
45
yes
yes
yes
yes
yes
yes
<5
3-4
<25
44
50
yes
yes
yes
3
5
26
50
no
11
48
43
no
9
20
40
no
no
12
9
14-20
35-50
47-54
no
13
28
48
7-8
41-48
45
yes
no
14
20
42
49
yes
no
12
38
42-46
yes
no
13
30
50
yes
no
14
23
51
41
no
11
Fahay (1983), Russell (1976), this study
Russell (1976), Schmidt (1906b), this study
Fish (1932), Jude (1982b), Snyder (1979)
Fahay (1983), Hardy (1978a), Russell (1976)
Demir (1982), Markle (1982)
D'Ancona (1933a), Fahay (1983), Russell (1976)
Fahay (1983), Hardy (1978a)
Russell (1976), this study
Russell (1976), Schmidt (1907b). this study
Russell (1976), Schmidt (1905a, 1906a), this study
Russell (1976), this study
Fahay (1983), Russell (1976), this study
Fahay (1983), Scott (1982)
Mukhacheva and Zviagina (1960), this study
Russell (1976), Schmidt (1905a, 1906a), this study
Matareseet al. (1981)
Dunn and Vinter(1984)
This study
Zviagina (1961)
Dunn and Vinter (1984), this study, (T. Nishiyama, pers.
comm. July 15, 1982)
Russell (1976), Schmidt (1905a), Seaton and Bailey (1971)
both between and within genera (Table 79, Fig. 147A-F). At
some size, phycine larvae usually have a single postanal pigment
bar located about midtrunk, but Ciliata has two bars (which
disappear during ontogeny) and Raniceps has none. Phycts lar-
vae less than 4.3 mm in length are not known, but larger post-
flexion larvae have a single midtrunk patch of pigment. The
location of the pigment bar varies among species of Gaidrop-
sarus. Postanal pigment spots along the ventral body midline
occur in Raniceps (anteriorly) and in Ciliata. Caudal pigment
can be present or absent and may be taxonomically significant
at the species level.
Phycines have two dorsal fins and one anal fin (Svetovidov,
1 948); the first and second dorsal fins are only slightly separated
(Table 81). A predorsal bone is present in Urophycis, Phycis
chestcri (two in P. blennoides) and Raniceps. but is wanting in
the other genera. X/Y bones are present (Fig. 148 A), or usually
present in Raniceps (Fig. 148B). Neural and haemal spines on
PU, are distally flattened except in Raniceps, in which those
bones are distally rounded. We detected evidence of ontogenetic
fusion of the hypural bones in Raniceps. as hypurals 2 and 3
are bifurcate distally (Fig. 148), and a sixth hypural bone was
found in one larva.
Subfamily Gadinae (Tables 77-82. Figs. 149-1 5 D. — lhis
subfamily contains the "true cods." There are approximately
30 nominal species presently assigned to twelve genera. They
are found in the North Atlantic, North Pacific and Arctic oceans
except for a single species, Micwmesistius australis. which is
distributed in the western South Atlantic Ocean to 60°S (Merrett,
1963;Shust, 1978).
The subfamily Gadinae is characterized as follows [MarkJe
(1982); this study]: egg diameter relatively large (0.9 to 1.9 mm);
no oil globule; vertebrae moderately numerous (39-64 total
vertebrae, 1 7-26 precaudal); pterotic spines absent; pelvic fin
ray formation at the same time as notochord flexion; .\ and v
bones absent; 4-5 primary caudal fin rays; 46-70 total caudal
fin rays; relatively few total dorsal and total anal fin elements
(D, 45-67 and A, 35-65).
Eggs of gadines are well known, with eggs of one or more
species of each genus described, except for Gadiculus and Arc-
togadus (Table 77). Most species shed small, planktonic eggs,
but demersal eggs are deposited by a number of species (Gadus
macrocephalus and both species of Microgadus. Eleginus. and,
presumably, Arctogadus). Characteristic pigment develops on
late stage embryos which aids in their identification.
Gadine larvae are also well known except for Arctogadus (Ta-
bles 78, 79). Length at hatching ranges from 2 to 6 mm, yolk
absorption (when known) occurs relatively early, notochbrd
flexion occurs from about 7 to 1 7 mm, and transformation to
the juvenile stage occurs at about 25-40 mm. The duration of
the pelagic state is moderate to long. Preflexion larvae typically
are moderately slender, tapering toward the tail, while flexion
larvae tend to be more robust (Table 78).
Head pigment is more diverse and diagnostically more im-
portant in this subfamily than in the lotines or phycines. Larvae
of most genera have pigment on the dorsal head and on the
mouth (usually the dentary). In some genera, the presence (e.g.,
Eleginus) or absence (e.g., Boreogadus) of gular and isthmus
pigment is important in identification. Absence of ventral gut
melanophores in certain size larvae (e.g., Boreogadus, Melano-
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 146. Caudal fin of Brosme hrosme. 45.2 mm SL. Hyl = Hypural bone 1; Hy2-3 = Hypural bones 2 and 3; Hy4-6 = Hypural bones 4,
5, and 6; EP, = Epural bone 1; EP, = Epural bone 2; U, = Ural centrum 1; U; = Ural centrum 2; PU, = Preural centrum 1; PU|„ = Preural
centrum 10. (Huntsman Mar. Lab., H 9742, stored at NWAFC).
grammus) or the presence and distribution of such pigment (e.g.,
Pollachius virens. P. pollachius) is also of diagnostic value (Rus-
sell, 1976; Matarese et al., 1981; Dunn and Vinter, 1984).
Pigment in the postanal region is also diverse and usually of
value in discriminating among species (e.g., Russell, 1976). For
purposes of discussion here we divide the gadines into three
groups based on their postanal pigment patterns: those genera
without postanal pigment bars in preflexion larvae, those genera
in which individual species may or may not possess such bars,
and those genera possessing one or two postanal pigment bars
(Table 79).
Merlangius and Melanogrammus lack postanal pigment bars
(Fig. 149C, D). In Merlangius. postanal pigment develops along
the dorsal body midline and extends to nearly three-quarters
the length of the body. Ventral pigment consists of a row of
melanophores from the anus to the caudal fin. Preflexion larvae
of Afelanogratnmus lack the dorsal line of pigment, but possess
the continuous ventral line.
Within Trisopterns and Pollachius, some species have one or
two postanal pigment bars, whereas others lack such bars [7'.
esmarkii. T. minutus. and P. pollachius (RusseW. 1976)]. In those
species possessing postanal bars, T. luscus (Fig. 149 A) has a
single pigment bar, with dorsal and ventral midline pigment
extending to about one-half the postanal body (Russell, 1976).
Pollachius virens (Fig. 149C) has two postanal pigment bars, the
anterior of which is close to the vent. Of those without pigment
bars, T. minutus and P. pollachius possess dorsal and ventral
lines of pigment extending to about three-quarters of the body
length; caudal peduncle pigment may be present in certain size
larvae of the former species, but is normally lacking in the latter.
Gadiculus has one postanal pigment bar located posterior to
the midpoint of the postanal region whereas Micromesistius has
a single bar near the midpoint of this region (Figs. 149F and
1 50E). The dorsal stripe is slightly longer than the ventral stripe.
Mediolateral pigment between the dorsal and ventral bars de-
velops during ontogeny, but the caudal peduncle area is not
pigmented. Gadus (in those species whose larvae are known),
Microgadus. Eleginus, Boreogadus, and Thcragra have two
postanal bars of pigment (not known for Arctogadus) as shown
in Figs. 149E and 150A-C. In some genera (e.g., Boreogadus)
the dorsal stripe of each bar is longer than the ventral stripe; in
others, the ventral stripe is longer than the dorsal (Gadus. Mi-
crogadus. Eleginus, and Theragra). The anterior end of the ven-
tral stripe may be near the anus (e.g., Gadus), or some distance
from It (e.g., Boreogadus). and the ventral stripes may be com-
posed of a single row of melanophores on each side of the body
midline (e.g., G. macrocephalus), a double row on each side of
the midline (E. gracilis), or on the ventral midline with scattered
pigment on each side of the body {B. saida). Caudal peduncle
pigment may be present or absent.
DU>fN AND MATARESE: GADIDAE
289
Table 79. Selected Pigmentation Characters Useful in Identifying Preflexion and Flexion Larvae of the Family Gadidae.
Postanal pigment
Bars Nuinber Dorsal Ventral
Flexion (any of stnpcs smpes
(mm) size) bars continuous continuous Description of pigment
Hypural
inargin
Pelvic
fins
Diagnostic
Lotinae
Brosme
Molva
Lota'
Phycinae
Enchelyopits
Gaidropsarus
Phycis
Urophycis
Ciliala
Raniceps
Gadinae
Tnsopterus
Merlangius
Poltachius
14-25 Yes
9-14
14-19
Yes
No
May
No
5-7 Yes 1
5-7 Yes 1
5-8 No?2
4-5 Yes
5-8 Yes
7-12 No
7-1 1 Yes/no
8-13 No
11-16 Yes/no
No
Yes
Yes Yes
(1 1 mm) (14 mm)
Melanogrammus 10-16 No
Within bars only
Above and
Yes
Caudal, 2 bars,
Fahay (1983),
below
pelvics
Russell
(1976)
May occur along
May occur
Yes
Caudal when
Russell (1976).
dorsal midline.
above
present, 2
Schmidt
mediolateral
and be-
bars, pelvics
(1906b,
within bars
low
1907a)
Along dorsal and
No
No
Lack of pig-
Fish (1932),
ventral midline.
ment in lar-
Hardy
mediolateral
vae <6 mm,
or dorsal/
ventral rows
(1978a),
Jude
(1982b),
Snyder
(1979)
Bar at mid-trunk
Yes
Yes
Mid-trunk bar,
pelvics
Fahay (1983),
Hardy
(1978a),
Russell
(1976)
Bar location vari-
May
Yes
Bar location.
Demir(1982),
able, dorsolat-
presence of
Markle
eral at bar
caudal
(1982)
Several spots
Yes
Yes
Lack of dorsal
D'Ancona
along ventral
initially, pel-
(1933), Fa-
midline
vic
hay (1983),
Hardy
(1978a),
Russell
(1976)
Variable, but mid-
May
Yes'
Postanal pat-
Fahay (1983),
trunk patch, me-
tern, pelvic
Hardy
diolateral at bar
tips
(1978a)
Bars disappear.
Yes
Yes
Loss of dorsal
Russell (1976),
ventral midline
pigment, pel-
this study
only
vics
Anterior ventral
No
Yes
LIpper body
Russell (1976),
midline, upper
pigment, pel-
Schmidt
body
vics
(1907b)
Variable length
No
No
Lack of bars
Russell (1976),
dorsal, ventral
(some
Schmidt
midline, medio-
species) re-
(1905a,
lateral within
duced dorsal
1906a)
bar
midline
Dorsal (shorter)
Yes
No
Length of dor-
Russell (1976)
and ventral
sal midline
midline dorso-
ventrolateral
Dorsal, and dou-
No
No
Unpigmented
Fahay (1983),
ble ventral mid-
posterior 1/4
Fridgeirsson
line, extends
body
(1978),
3/4, some
Hardy
mediolateral
(1978a),
Russell
(1976)
Small, double
Yes
No
Ventral mid-
Fahay (1983),
ventral midline
line, nape
Hardy
(1978a),
Russell
(1976)
290
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
Table 79. Continued.
Postanal pigment
Flexion
(mm)
Bars
(any
size)
Number
of
bars
Dorsal
stnpes
contmuous
Ventral
stnpes
continuous
Descnption of pigment
Hypural
margin
Pelvic
tins
Diagnostic
Gadus
10-17
Yes
Yes
{6 mm)
Yes
(6 mm)
Initially posterior
stripes longer,
mediolateral
Gadiculus
7-13
Yes
1
Bar posterior to
mid-trunk
Microgadm
8-15
Yes
2
Yes
(15 mm)
Yes
(6 mm)
Bars anterior, me-
diolateral
Eleginus
11-17
Yes
2
Yes
(10 mm)
Yes
(7 mm)
Ventral stripes
longer, double
ventral row
each side of
midline, medio-
lateral
Boreogadus
11-17
Yes
2
Yes
Yes
Ventral stripe
(7 mm) ( 10 mm)
shorter, medio-
lateral
Arctogadus
Dorsal, ventrolat-
eral margins,
mediolateral
Theragra
10-17
Yes
2
Yes
(13 mm)
May
Posterior ventral
stripe longer
than dorsal, sin-
gle ventral row
each side mid-
line, medio-
lateral
Mkromesislius
8-13
Yes
1
Bar mid-trunk,
dorsal stnpe
longer, nape
May
No
No
Yes
No
No
No
No
No
No
May
No
No
No
Continuous
Dunn and
stnpes, ven-
Vinter
tral gut
(1984), Ma-
tarese et al.
(1981), Mu-
khacheva
and Zvi-
agina(1960)
Schmidt
(1906a)
Posterior loca-
Russell (1976)
tion of bar
Schmidt
(1905a,
1906a)
Bar location.
Matarese et al.
ventral gut.
(1981)
caudal
Continuous
Dunn and
stripes.
Vinter
mediolateral,
(1984)
ventral gut
Bars, medio-
Dunn and
lateral
Vinter
(1984),
this study
Zviagina
(1961)
Bar location.
Dunn and
length
Vinter
(1984), Gor-
bunova
(1954), Ma-
tarese et al.
(1981)
Bar location,
Lisovenko et
upper body
al. (1982),
Russell
(1976),
Schmidt
(1905a),
Weiss
(1974)
' See lext for discussion of pigmenlation in Lota
' In Phycines. Ihe bar occurs early. Specimens of Ph\
' Pelvic fins not pigmented in Li regia.
/.s 3-4 mm not available to us.
Gadines have three dorsal fins and two anal fins. The distances
(number of intemeural bones) between dorsal fins 2 and 3 and
between anal fins 1 and 2 vary among genera. In all genera, the
dorsal and anal fins are separate from the caudal fin. The lower
branch of the posttemporal tends to become more elongate than
in lotines or phycines (longest in Arctogadus and Micromesis-
tius), the postcieithntm always has an expanded distal head, and
the posterior process of the basipterygia tends to be short or
even lacking. Predorsal bones are absent in Gadinae. The first
and second dorsal fins are not usually internally continuous, but
the second and third dorsal fins are always internally continuous.
The neural and haemal spines on PU, are distally flattened in
all gadine genera and only three hypural bones (including the
parhypural) are present (Table 82, Fig. 151).
Fig. 147. (A) Preflexion larva of Enchelyopus cimbrius. 3.7 mm SL (Huntsman Mar. Lab., H-5388, stored at NWAFC); (B) Flexion larva of
Gaidropsarus mediterraneus, 6.1 mm SL (from Demir, 1982); (C) Flexion larva of Phycis blennoides. 4.3 mm SL (from Russell, 1976); (D)
Flexion larvae of Urophycis sp., 4.2 mm SL (Huntsman Mar. Lab., H- 16384, stored at NWAFC); (E) Flexion larva of Ciliata sp., 4.4 mm SL
(Zool. Mus. Copenhagen, stored at NWAFC); and (F) Preflexion larva of Raniceps raninus, 4.7 mm SL (Inst. Sci. Tech. Peches. Marit. Nantes,
stored at NWAFC).
DUNN AND MATARESE: GADIDAE
291
B
292
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
^e- Hy 4-6
Fig. 148. (A) Caudal fin oi Phyas blennoides. 82.3 mm SL (British Mus. Nat. Hist. 1976. 7.30.110-119); Hyl = Hypural bone 1; Hy2-3 =
Hypural bones 2 and 3; Hy4-6 = Hypural bones 4, 5, and 6; EP, = Epural bone 1, EP, = Epural bone 2; X = x bone, Y = y bone; U, = Ural
centrum 1; LI, = Ural centrum 2; PU, = Preural centrum 1; PU^ = Preural centrum 6; (B) Caudal fin oi Raniceps raninus. 44.4 mm SL (British
Mus. Nat. Hist. 1971.2-16.640); symbols as m (A).
Fig. 149. (A) Flexion larva of Trisopterus luscus. 7.5 mm SL (Inst. Sci. Tech. Peches Marit., Nantes, stored at NWAFC); (B) Preflexion larva
of Merlangius merlangus. 5.0 mm SL (Inst. Sci. Tech. Peches Mant., Nantes, stored at NWAFC); (C) Preflexion larva of Pollachius virens. 5.9
mm SL (Huntsman Mar. Lab., H-8057, stored at NWAFC); (D) Preflexion larva of Melanogrammus acglefinus. 6.1 mm SL (Huntsman Mar.
Lab., H-9473, stored at NWAFC); (E) Preflexion larva of Gadus macrocephalus, 4.4 mm SL (from Dunn and Vinter, 1984); and (F) Preflexion
larva of Gadiculus argenteus. 3.7 mm SL (Zool. Mus. Copenhagen, stored at NWAFC).
DUNN AND MATARESE: GADIDAE
293
B
294
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 80.
Summary of Osteological Characters of the Pectoral and Pelvic Girdles, Axial Skeleton, and Median Fins in Pre-
TRANSFORMATION LaRVAE OF REPRESENTATIVES OF THE GeNERA OF THE FAMILY GaDIDAE.
Relative length of lower
fork of posttemporal
Shape of
poslcleithrum
Length of postenor process
of bastpterygia
Number
pre- Number
dorsal dorsal
bones fins
Lotinae
Brosme brosme
Molva dipterygia
Lola Iota'
Phycinae
Enchelyopus cimbrius
Gaidropsarus sp.
Phycis blennoides'
Vrophycis sp.
Ciliata must el la
Raniceps raninus
Gadinae
Trisopterus liiscus
Merlangius merlangus
Pollachius virens
Melanogrammus
aeglefinus
Gadus macrocephalus
Gadiculus argenteus
Microgadus proximus
Eleginus gracilis
Boreogadus saida
Arctogadus horisovi'
Theragra chalcogramma
Micromesislius
poutassou
Very short, less than '/,o length
of upper fork
Short, less than 'A length of
upper fork
Short, less than % length of
upper fork
Short, less than '/j length of
upper fork
Short, less than % length of
upper fork
Short, less than V, length of
upper fork
Short, less than % length of
upper fork
Very short, less than y,,, length
of upper fork
Short, less than '/, length of
upper fork
Moderately long, about Vi
length of upper fork
Moderately long, about %
length of upper fork
Moderately long, about V^
length of upper fork
Moderately long, about %
length of upper fork
Moderately long, about '/,
length of upper fork
Long, about % length of upper
fork
Long, about % length of upper
fork
Moderately long, about %
length of upper fork
Long, about 'A length of upper
fork
Long, about -/, length of upper
fork
Long, about y, length of upper
fork
Long, about -/, length of upper
fork
Long, thin, slightly curved,
distal head slightly expand-
ed
Long, thin, slightly curved,
distal head slightly expand-
ed
Long, thin, slightly curved,
distal head wanting
Moderately long, thin, not
curved, no distal head
Long, thin, not curved, no
distal head
Long, thin, slightly curved, no
distal head
Long, thin, curved, no distal
head
Short, thin, no distal head
Lower bone long, thin, point-
ed; upper bone short, ob-
long
Long, thin, slightly curved,
slightly expanded head
Long, thin, straight, expanded
distal head
Long, thin, curved, with ex-
panded distal head
Long, thin, slightly curved,
slightly expanded distal
head
Long, thin, strongly curved,
expanded distal head
Short, relatively wide, pointed
distally, slightly expanded
head
Long, straight, recurved dis-
tally, with expanded head
Long, straight, recurved dis-
tally, with expanded head
Long, thin, slightly recurved
distally, moderately ex-
panded head
Long, thin, striaght, expanded
head
Long, thin, recurved, expand-
ed distal head
Long, thin, recurved, expand-
ed distal head
Long, about -/j length anterior 0
process
Moderate, about Vi length an- 0
terior process
Short, about '/, length anterior 0
process
Very long, about 2 x length 0
anterior process
Long, about -/, length anterior 0
process
Very long, about 1 '/. x length 2
anterior process
Very long, about 4 x length 1
anterior process
Very long, about I 'A x length 1
anterior process
Long, about 74 length anterior 1
process
Absent
Very short, less than '/lo
length anterior process
Very short, less than '^j 0
length anterior process
Absent 0
0
Moderate, about '/; length an- 0
terior process
Short, about '/, length antenor 0
process
Very short, about '/,o length 0
anterior process
Short, about % length anterior 0
process
Short, about % length anterior 0
process
Short, about % length ante- 0
rior process
Short, about '/i length anterior 0
process
Absent 0
' Juvenile specimens only examined.
Comments on the Systematic Relationships
OF Subfamilies in the
Family Gadidae
Gadoid fishes comprise a coinplex and rather confusing array
of teleosts possessing both relatively primitive and apparently
derived character states (Cohen, this volume; Fahay and Markle,
this volume). In our analyses of character states we generally
follow Markle (1982), Fahay and Markle (this volume) and
Cohen (this volume). For outgroup comparisons, we have ex-
amined the osteology of representatives of a limited array (14
families) of gadiform and non-gadiform fishes whose utility is
limited because we lack, in many cases, ontogenetic series. We
contrast here the characters of the three recognized subfamilies
(Markle, 1 982) with Merluccius. insofar as possible, as the genus
is variously considered primitive in the gadid-merlucciid lineage
(Danil'chenko, 1947, 1950; Rosen and Patterson, 1969; Cohen,
Table 81.
Summary of Osteological Characters of the Median Fins in Pretransformation Larvae of Representatives of the Genera
OF THE Family Gadidae.
Relative dislance between
dorsal fins one and two
Relative distance between
dorsal fins two and three
Relative distance between
postenormost dorsal and
caudal fin
Number
of anal Relative distance between
fins anal fins one and two
Lotinae
Brosme brosme
Molva dipten'gia
Lota lota'
Phycinae
Enchelyopus cimbrius
Gaidropsarus sp.
Phycis blennoides'
Urophycis sp.
Ciliala muslella
Raniceps raninus
Gadinae
Tnsopterus luscus
Merlangius merlangus
Pollachius virens
Melanogra mmiis
aeglefinus
Gadus macrocephatus
Gadiculus argenteus
Microgadus proximus
Eleginus gracilis
Boregadus saida
Arclogadus borisovi'
Theragra
chalcogramma
Micromesislius
poutassou
Very close, about '/, in-
temeural space; no in-
temeural bones
Very close, about '/, in-
teraeural space; 1 or 0
intemeural bones
Wide, about 2 inter-
neural spaces; no in-
temeural bones
Very close, about 'A in-
temeural space; no in-
temeural bones
Very close, about '/j in-
temeural space; 1 in-
temeural bone
Nearly continuous; I in-
temeural bone
Very close, about '/, in-
temeural space; I in-
temeural bone
Close, about 1 inter-
neural space; I re-
duced intemeural
bone
Close. '/;-l intemeural
space; 0 or 1 inter-
neural bones
Close, usually '/,-l inter-
neural space; 2 inter-
neural bones
Close, about 'A inter-
neural space; no inter-
neural bones
Close, about '/, inter-
neural space; 0-2 in-
temeural bones
Close, about '/, inter-
neural space; 0 or 1
intemeural bones
Moderately wide, about
I intemeural space;
no intemeural bones
Close, about '/, inter-
neural space; 0 or I
intemeural bones
Moderately wide, about
I '/, intemeural spaces;
0 or 1 intemeural
bones
Wide, about 2 inter-
neural spaces; 0 or I
intemeural bones
Wide, about 2'/, inter-
neural spaces; no in-
temeural bones
Moderately wide, about
1 Vj intemeural spaces;
0 or I intemeural
bones
Wide, about 2 inter-
neural spaces; no in-
temeural bones
Very close together, 0-2
intemeural bones
Close together. 2-4 in-
temeural bones
Close together, 3-4 in-
temeural bones
Close together, 2-3 in-
temeural bones
Close together, 2-3 in-
temeural bones
Moderately wide, about
5-6 intemeural bones
Moderately wide, 4-7
intemeural bones
Moderately wide, 5-7
intemeural bones
Wide, 6-7 intemeural
bones
Wide, 5-8 intemeural
bones
Wide. 7-9 intemeural
bones
Very wide, 20-2 1 inter-
neural bones
Close, about 1 inter-
neural space
Close, about I 'A inter-
neural spaces
Close, about 1 inter-
neural space
Close, about 1 inter-
neural space
Close, about 1 inter-
neural space
Close, about I inter-
neural space
Close, about 1 inter-
neural space
Close, about 1 inter-
neural space
Close, about 1 inter-
neural space
Wide, 2-2 'A intemeural
spaces
Wide. 2-2', intemeural
spaces
Wide, about 2 inter-
neural spaces
Wide, about 2 inter-
neural spaces
Wide, about 3 inter-
neural spaces
Close, about 1 'A inter-
neural spaces
Wide, about 3-3'/2 inter-
neural spaces
Wide, about 3 inter-
neural spaces
Wide, about 3 inter-
neural spaces
Wide, about 4 inter-
neural spaces
Wide, about 3-3'/, inter-
neural spaces
Wide, about 2'/, inter-
neural spaces
Very close, from 0-2
intemeural bones
Very close, from 0-2
intemeural bones
Close, from 2-3 inter-
neural bones
Close, usually 2 inter-
neural bones
Moderately wide, 4 or
5 intemeural bones
Moderately wide, 4 or
5 intemeural bones
Moderately wide, about
4 intemeural bones
Wide, 6 or 7 intemeu-
ral bones
Wide. 5-7 intemeural
bones
Wide, 5 or 6 intemeu-
ral bones
Wide, 4-6 intemeural
bones
2 Very close, 1 intemeu-
ral bone
' Juvenile specimens only examined.
296
ONTOGENY AND SYSTEMATICS OF FISHES- AH LSTROM SYMPOSIUM
B
D
DUNN AND MATARESE: GADIDAE
297
Table 82. Summary of Osteological Characters of the Caudal Fin in Pretransformation Larvae of Representatives of the Genera
OF the Family Gadidae.
Taxon
Relalive distance between postenor
margin of anal to caudal
Shape neural spine on
preural centrum one
Shape haemal spine on
preural centrum one
X. Y bones
Lotinae
Brosme brosme
Very close, about '/, intemeural
Distally flattened
Rounded distally
Absent
Molvu dipterygia
space
Moderate, about 2 intemeural
Distally flattened
Rounded distally
Usually absent
Lola tola'
spaces
Close, about 1 intemeural
space
Distally flattened
Distal -/j slightly flat-
tened
Usually absent
Phycinae
Enchelyopus cimhhus
Close, about 1 intemeural
Distally flattened
Distally flattened
Present
Gaidropsarus sp.
space
Close, about 1 intemeural
Distally flattened
Distally flattened
Present
Phycis blennoides'
space
Close, about 1 intemeural
Distally flattened
Distally flattened
Present
Urophyas sp.
space
Close, about 1 intemeural
Distally flattened
Distally flattened
Present
Ciliala mustelta
space
Close, about 1 intemeural
Distally flattened
Distally flattened
Present
Ramceps raninus
space
Close, about 1 intemeural
space
Distal '/, rounded
Distal ■/, rounded
Usually pres-
ent
Gadinae
Trisoplerus luscus
Wide, about 2-2'/, intemeural
Distally flattened
Distally flattened
Absent
Merlangius merlangus
spaces
Moderate, about 1 '/,-2 inter-
Distally flattened
Distally flattened
Absent
Pol lac hi us virens
neural spaces
Wide, about 2-2'/, intemeural
Distally flattened
Distally flattened
Absent
Melanogrammus
aeglefinus
Gadus macrocephalus
spaces
Moderate, about 2 intemeural
spaces
Wide, about 3 intemeural
Distally flattened
Distally flattened
Distally flattened
Distally flattened
Absent
Absent
Gadiculus argenleus
spaces
Close, about 1 '/, intemeural
Distally flattened
Distally flattened
Absent
Microgadus proximus
spaces
Close, about 1 '/, intemeural
Distally flattened
Distally flattened
Absent
Eleginus gracilis
spaces
Wide, about 3 intemeural
Distally flattened
Distally flattened
Absent
Boreogadus saida
spaces
Wide, about 3 intemeural
Distally flattened
Distally flattened
Absent
Arctogadus borisovi'
spaces
Wide, about 3 intemeural
Distally flattened
Distally flattened
Absent
Theragra chalcogramma
spaces
Wide, about 3 intemeural
Distally flattened
Distally flattened
Absent
Micromesistius
spaces
Moderate, about 2 intemeural
Distally flattened
Distally flattened
Absent
poutassou
spaces
' Juvenile specimens only examined.
this volume), a basal gadid (Mujib, 1 967), a medial gadid related
to gadines (Svetovidov, 1948, 1969) or of questionable rela-
tionship (Fahay and Markle, this volume). We present here our
interpretation of the relationships of subfamilies of gadid fishes.
Egg diameter is largest in lotines and gadines, smallest in
phycines (Table 77). A single oil globule is present in lotines;
multiple oil globules, which duiing development coalesce into
one, are found in most phycines (not yet reported to occur in
Ramceps). and are absent in gadines. Merlucctus has a mod-
erately sized egg (0.8-1.2 mm) with a single oil globule (Ahl-
strom and Counts, 1955; Russell, 1976; Fahay, 1983). Markle
( 1982) considered small ( < 1 mm) eggs, possessing an oil globule,
the primitive state. We agree, but also consider multiple oil
globules, which coalesce into one the most primitive state.
Fig, 150. (A) Preflexion larva of Microgadus proximus. 3.6 mm SL (from Matarese et al., 1981); (B) Preflexion larva of Eleginus gracilis, 5.0
mm SL (from Dunn and Vinter, 1984); (C) Preflexion larva of Boreogadus saida. 6.3 mm SL (from Dunn and Vinter, 1984); (D) Preflexion larva
of Theragra chalcogramma. 6.2 mm SL (from Matarese et al., 1981); and (E) Flexion larva of Micromesistius poulassou. 8.0 mm SL (Zool. Mus.
Copenhagen, stored at NWAFC).
298
ONTOGENY AND SYSTEM ATICS OF FISHES- AHLSTROM SYMPOSIUM
Hy4-6
Fig. 151. Caudal fin o( Microgadus proximus, 41.1 mm SL. Hyl = Hypural bone 1; Hy2-3 = Hypural bones 2 and 3; Hy4-6 = Hypural bones
4, 5, and 6; EP, = Epural bone 1; EP, = Epural bone 2; U, = Ural centrum 1; U, = Ural centrum 2; PU, = Preural centrum 1; PUk, = Preural
centrum 10 (after Matarese et al., 1981).
Lotinae lai-vae are relatively elongate and somewhat narrow
at the pectoral fin base; the former state is partially due to their
numerous vertebrae (Table 78). In contrast, phycines are shorter
and stockier in appearance, deep bodied at the pectoral fin base,
and morphologically somewhat resemble scorpaeniform larvae.
Ramceps larvae are morphologically the most divergent, ap-
pearing tadpole shaped due to their depth at the pectoral fin
base. Gadines are somewhat shorter in appearance, and deeper
bodied, than lotines, but morphologically intermediate between
phycines and lotines. Merluccius larvae are similar to gadines
in overall shape (Fig. 143D in Fahay and Markle, this volume).
Length at hatching is smallest in most phycines and somewhat
larger in Raniceps. lotines, and gadines (Table 78). Notochord
flexion occurs at quite small sizes in phycines (except in Ran-
iceps), relatively larger sizes in lotines, and intermediate sizes
in gadines. A silvery prejuvenile stage is present in phycines
(not recorded for Raniceps), but a pelagic stage of varying du-
ration (Table 78) is probably present in all gadid larvae (Fahay
and Markle, this volume). Merlucciiis hatches at moderate lengths
(2.6-3.8 mm; Ahlstrom and Counts, 1955; Russell, 1976; Fahay,
1983), notochord flexion begins at about 9 mm, and transfor-
mation begins at 20-25 mm, somewhat similar to gadines. Elon-
gate pelvic fins develop precociously in all lotines (except Lota)
and phycines, but not in gadines. Pelvic fins in Merluccius are
shorter than in phycines. but longer than in gadines. Fahay and
Markle (this volume) noted similarities in fin development be-
tween Merluccius and gadines in that the caudal develops first.
In gadines, however, the pelvic fins develop last. In Merluccius.
it is the second fin to develop.
Pigment patterns are shared by Brosme and Molva. but not
Lota, whose pigment resembles certain gadines (e.g., Pollachius
pollachius, Trisopterus minutus). Two kinds of pigment patterns
have been identified for phycines: either dorsal saddles of pig-
ment in the postanal region (Enchelyopus cimbrius, Gaidrop-
sarus mediterraneus. Urophycis chuss) or a ventral series of me-
lanophores (Phycis blennoides, Ciliata. and Raniceps). Gadines
have either one or two postanal bars or dorsal and ventral lines
of pigment. In gadines the pelvic fins lack pigment such as that
present in lotines and phycines. In comparison, Merluccius has,
in certain species (e.g., M. product us. Ahlstrom and Counts,
1955), a single postanal band of pigment, but two in M. albidus
and M. hilinearis (Fahay, 1983) and three lateral melanophores
(M. merluccius. Russell, 1976). Pelvic fins are pigmented in
some Merluccius.
Lotines have one (Brosme) or two (Moha. Lota) dorsal fins;
when two fins are present, they are internally continuous (Mar-
kle, 1982). Phycines have two dorsal fins, the first specialized
(Cohen and Russo, 1979), and are internally continuous (Mar-
kle. 1982). Gadines have three dorsal fins, of which the second
and third are always internally continuous. The posterior mar-
gins of the dorsal and the anal fins are close to the procurrent
rays of the caudal fin in lotines and phycines, whereas in gadines
these fins are generally some distance from the caudal fin, as is
the case in Merluccius (Fahay and Markle, this volume).
Certain trends in osteological structures can be noted in the
family Gadidae. Transient pterotic spines are present in some
phycines (some f/!i'ai. Gaidropsarus. Markle, 1 982) and Ciliata
(this study), and are lacking (so far as is known) in other phy-
cines, lotines, gadines, and in Merluccius. The distribution of
branchiostegal rays varies among gadid genera. The seven bran-
chiostegal rays in Brosme are carried on the outer surface of the
ceratohyal, whereas in Gadus and Lota (Mujib, 1967, 1969). as
well as in Theragra (Dunn and Vinter, MS), the three anterior
rays are internal and the posterior four are external. Raniceps
has one branchiostegal ray on the epihyal, a character considered
primitive and shared (so far as known) with Urophycis chuss
and Merluccius (Mujib, 1967; Inada, 1981b), whereas gadines
have all seven branchiostegal rays on the ceratohyal (this study).
DUNN AND MATARESE: GADIDAE
299
Phycinae
Gadinae
Raniceps
/
Fig. 152. Proposed relationships of gadid subfamilies.
The ventral branch of the posttemporal is shortest in lotines
and phycines (moderately long in Merluccius) and longest in
gadines. The gadids we examined all have four pectoral radials,
except for a single specimen of Ciliata with five radials on one
side. Phycines lack an expanded distal head on the postcleithra.
Raniceps. however, has two postcleithra; the upper is oblong in
shape, the lower is distally pointed as in phycines. Among the
lotines, the distal end of the postcleithrum is slightly expanded
in Brosme and Molva whereas the postcleithrum is distally
pointed in Lota. In Merluccius it is moderately expanded while
in gadines, the postcleithrum is considerably expanded at its
tip, which we infer is a derived condition. Predorsal bones are
present in some, but not all, phycines (Urophycis. Phycis. and
Raniceps). but are lacking in lotines, gadines, and Merluccius:
the loss is considered an advanced state. The posterior process
of the basipterygia is quite long in some phycines (and in Mer-
luccius), moderately long in lotines, and shortest in gadines, and
the latter state is considered derived.
The shape of the neural and haemal spine on PU, varies
among genera. The neural spine in Raniceps is distally rounded
(a primitive condition), but this spine is flattened in all other
gadids, as it is in Merluccius. Raniceps. Brosme, and Molva
have a rounded haemal spine on PU,, in contrast to the flattened
tip on all other gadids (and Merluccius): x/y bones are present
in all phycines (usually present in Raniceps) and Merluccius,
but are absent {Brosme) or usually absent {Molva, Lota) in
lotines (Markle, 1982; this study) and are absent in gadines. All
gadids and Merluccius (Ahlstrom and Counts, 1955; Inada.
1981b) have three hypural bones (including the parhypural);
Raniceps alone, among the gadids examined by us, showed evi-
dence of ontogenetic reduction by fusion from six hypural bones
to three. As noted by Markle ( 1 982) and Fahay and Markle (this
volume), lotines and gadines have four or five primary caudal
fin rays, while phycines have five or six such rays. Merluccius
and Raniceps each have six primary rays (Inada, 1981b; this
study).
We consider Raniceps a basal gadid considering the following
characters: eggs small with a single oil globule; larvae tadpole-
or liparid-shaped; one branchiostegal ray on the epihyal; two
postcleithra present; a predorsal bone present; the neural and
haemal spines on PU, distally rounded; six hypural bones which
fuse into three during ontogeny; x/y bones usually present; and
six primary rays on the superior hypural bone.
We further consider phycines to be a more primitive group
than lotines based on the following characters: eggs small, with
multiple oil globules which coalesce into one during develop-
ment; larvae stocky and deep bodied (at the pectoral fin base);
elongate and precocious pelvic fins present; postcleithrum with-
out an expanded head; one or more predorsal bones present;
elongate pelvic process; and x/y bones present. Until the pres-
ence or absence of transient pterotic spines is established in all
phycine larvae, the most parsimonious explanation is that their
presence represents a derived character state.
Lotines, as presently constituted, appear to us to possess a
number of primitive and intermediate characters, as well as
some rather specialized traits: eggs moderately large with a single
oil globule; larvae elongate, relatively shallow at the pectoral fin
base; pelvic fins precocious, elongate and with the posterior
process of the basipterygium moderately long; postcleithnam
with slightly expanded head; predorsal bones absent; x/y bones
usually absent; and three hypural bones present. Brosme has
both apparently primitive (e.g., all branchiostegal rays carried
on the outside surface of the ceratohyal and a rounded haemal
spine on PU,) and derived characters (e.g., x/y bones always
lacking); its single dorsal fin was considered primitive by Sve-
tovidov (1948) or derived (within Lotinae, sensu Svetovidov,
1948) by Mujib (1969). As noted by Markle (1982), high total
dorsal and anal fin ray counts may be primitive for the order
Gadiformes.
Gadines seem to us a relatively homogenous group, charac-
terized by reductive (or lost) and apparently derived characters.
The former include: eggs without an oil globule; posterior pro-
cess of the pelvic bone reduced in length or wanting; predorsal
and x/y bones absent; and three hypural bones present. The
latter characters include: eggs moderate in size; larvae morpho-
logically uniform in appearance; lower branch of posttemporal
relatively long; postcleithrum with expanded head; and three
dorsal and two anal fins present, with the anal fins and dorsal
fins two and three internally continuous.
Our hypothesis of relationships of gadid subfamilies is pre-
sented in Figure 152. The relationships of a number of genera,
such as Brosme and Raniceps. and the relationships of Phycis,
Gaidropsarus, and Ciliata to other phycines still remain con-
fused. Based on early life history characteristics and osteology,
we consider Merluccius a gadid related to, but more primitive
than. Gadinae and, following Svetovidov (1948, 1969), restrict
Merlucciinae to this genus. The relationship of all nominal gadid
subfamilies requires further study.
Northwest and Alaska Fisheries Center, 2725 Montlake
Boulevard East, Seattle, Washington 981 15.
Bregmacerotidae: Development and Relationships
E. D. HouDE
THE codlets are small, gadiform fishes of pelagic habit found
in neritic and oceanic water of tropical and subtropical
seas. The family Bregmacerotidae (Gill, 1872) includes the single
genus Bregmaceros (Thompson, 1840), in which there are sev-
eral species. In recent reviews six (Belyanina, 1974) or seven
(D'Ancona and Cavinato, 1965) valid species have been rec-
ognized. The systematics remain confused, although Belyanina
(1974) has partly clarified species relationships. Larvae often
are among the ten most common families occurring in both
oceanic and coastal ichthyoplankton surveys in subtropical and
tropical waters (e.g., Ahlstrom, 1971; Moser et al., 1973; Houde
etal., 1979; Loeb, 1979; Richards, 1981). The species are mor-
phologically similar but most have distinctive meristics, from
which specific identifications usually are possible. Differences
in vertebral number and median fin ray counts serve to distin-
guish larval to adult stages while pigmentation differences and
the size at appearance of the single, first dorsal fin ray serve to
identify small larvae. Larval characters, particularly those of the
smallest individuals (1.5-3.0 mm SL), often are the best char-
acters for identification purposes. A careful examination of on-
togenetic evidence indicates that some species are still unde-
scribed and that misidentified Bregmaceros frequently have been
reported in the literature. Based on evidence from larval char-
acteristics there may be ten or more valid species in the world
oceans.
Species distributions. — Larvae of Bregmaceros commonly occur
between latitudes 40°S and 40°N (Table 83). D'Ancona and
Cavinato (1965) and, more recently, Belyanina (1974), have
reviewed distribution data on the known species. Centers of
abundance have been observed in the western Indo-Pacific and
Indian Oceans (Munro, 1950; D'Ancona and Cavinato, 1965;
Kotthaus, 1969; Belyanina, 1974), in the eastern Pacific (Ahl-
strom. 1971; Belyanina, 1974) and in the Caribbean Sea and
Gulf of Mexico (Belyanina and Lopes, 1974; Milliken, 1975;
Belyanina, 1980; Houde, 1981). Bregmaceros macclellandi is
circumtropical with areas of apparent high abundances in the
Caribbean Sea, western Indian Ocean and Indo-Malayan region.
It also occurs in the eastern Pacific. Bregmaceros atlanticus.
including the closely related Pacific Ocean form B. japonicus
(D'Ancona and Cavinato, 1965) also is circumtropical with an
apparent center of abundance in the western Atlantic. ' The latter
sometimes occurs in neritic waters. Several neritic species are
known, including B. nectahanus, B. arahicus, B. rarisquamosus,
B. bathymaster, B. caw/or; (Milliken and Houde, 1984) and the
Type A larva described by Houde (1981).
Neritic species vary in the breadth of their distributions. It
now seems certain that the Indo-Pacific B. nectabanus does not
occur in the western Atlantic and its occurrence in the eastern
Atlantic Ocean is uncertain. The species B. cantori. described
by Milliken and Houde (1984), is the most common bregma-
cerotid in the western Atlantic. It occurs in the Caribbean Sea
and Gulf of Mexico (Milliken, 1975; Houde, 1981), in the south-
west Atlantic Ocean off Brazil- and along the East Coast of the
United States.' The common bregmacerotid in the Gulf of Ca-
riaco, initially referred to as B. atlanticus (Mead, 1963) and
subsequently as B. nectabanus (Baird et al., 1973, 1974; Bely-
anina and Lopes, 1974) and that referred to as B. nectabanus
from the Caribbean Sea and Gulf of Mexico (Belyanina, 1980)
is B. canton (M\\\\ken, 1975; Houde, 1981; Milliken and Houde,
1984). Bregmaceros bathymaster has been collected only in the
eastern Pacific. It is abundant in the Gulf of Panama (D'Ancona
and Cavinato, 1965) and in the Gulf of California (Moser et al.,
1973). Bregmaceros rarisquamosus occurs in the Indian Ocean,
Bay of Bengal, Arabian Sea and western Pacific Ocean. It also
is present in the Persian Gulf'' where it occurs with B. necta-
banus and B. arabicus. Previously, B. arabicus had been re-
ported from the Arabian Sea, Bay of Bengal and East China Sea.
Larvae of an undescribed species, B. Type A, have been collected
in the western North Atlantic (Houde, 1981)' ^
Bregmacerotids reportedly occur from the surface to depths
of approximately 4,000 m, but are most common in the upper
300 m. Larvae generally occur from surface to 600 m depth,
neritic species tending to be closer to the surface than oceanic
species (D'Ancona and Cavinato, 1965). Some reported catches
from great depths may be in error. Adults and subadults of some
Bregmaceros undertake extensive vertical migrations and one
species (B. cantori) inhabits anoxic water during a part of the
day (Mead, 1963; Wilson. 1972; Baird et al., 1973; Milliken,
1975).
Family characteristics. — CharaclcTs defining Bregmacerotidae
were summarized briefly by Nelson ( 1 976) and more extensively
by D'Ancona and Cavinato (1965) and by Belyanina (1974).
Fahay and Markle (this volume) have tabulated meristic data
and discussed ontogenetic characters of Gadiformes, including
' Late larvae and juveniles that I examined from the eastern Pacific
appeared to be typical B. atlanticus but small larvae, which may have
been younger specimens of this species, did not resemble typical B.
atlanticus from the Atlantic. The eastern Pacific specimens were less
pigmented, with a prominent melanophore on the ventral midline, be-
tween the anus and the lip of the tail. Specimens were provided by Dr.
H. G. Moser, Southwest Fisheries Center, National Marine Fisheries
Service, La Jolla, California.
- I examined specimens of B. cantori from coastal waters of Brazil,
collected from latitudes of 22°S to 27°S, provided to me by Dr. Y.
Matsuura, Instituto Oceanografico, Universidade de Sao Paulo, Brazil.
' 1 examined specimens from R/V DOLPHIN cruises, taken from
Florida to the Carolinas, provided to me by M. P. Fahay, Sandy Hook
Laboratory, National Marine Fisheries Service, Highlands, New Jersey.
■■Houde. E. D., J. C. Leak, S. Al-Matar, and C. E. Dowd. 1981.
Ichthyoplankton abundance and diversity in the weslem Arabian Gulf.
Kuwait Institute for Scientific Research, Mariculture and Fisheries De-
partment, Final Report, Project MB- 16, 3 volumes. (This report was
not available for distribution at the time the present paper was written.)
' The Type A larva was present in collections from two R/V AL-
BATROSS cruises into the Caribbean Sea. 1 examined larvae provided
by Dr. W. J. Richards, Southeast Fisheries Center, National Marine
Fisheries Services, Miami, Florida.
300
HOUDE: BREGMACEROTIDAE
301
Bregmacerotidae. Bregmacerotids are small fishes, the largest
species, B. macclellandi, rarely exceeding 120 mm SL. They
have two dorsal fins, the first a single, elongate ray on the occiput.
The second dorsal fin and the anal fin are long with median rays
much reduced, giving the fins a divided appearance. The caudal
fin is separated from the dorsal and anal fins. Pelvic fins are
j ugular and consist of 5 (usually )-7 rays, the outer three elongate.
The olfactory nerves pass through a broad canal, wider than
that in Gadidae. The sacculus is very large. The swimbladder
does not contact the auditory capsules. There are a few pyloric
caeca. The vomer is toothed. A lateral line is present under
the second dorsal fin. Chin barbels are absent.
Development
Spawning. — Size at maturity is variable but generally <30 mm.
In one species, B. rarisquamosus. maturity is attained at < 1 5
mm (D'Ancona and Cavinato, 1965). Larvae occur in the tropics
and subtropics dunng all months, indicating protracted spawn-
ing, although seasonality is apparent for individual species in
some areas.
£■^^5. — Eggs are presumed to be pelagic. Excepting a single re-
port, the fertilized eggs and embryos of Bregmaceros species
have not been described. Pertseva-Ostroumova and Rass (1973)
described fertilized eggs, attributable to B. atlaniicus. as pelagic
with smooth chorion, small perivitelline space and homoge-
neous yolk containing an oil globule. They reported the egg
diameter to be 1.1 mm and the oil globule diameter to be 0.20
mm. In my opinion, it is unlikely that Bregmaceros eggs are
that large because newly-hatched larvae are only 1.5 mm long.
Ahlstrom's'' unpublished notes give diameters of Z?. bathyniaster
eggs as 0.84-1.00 mm and indicate that a single oil globule is
present.
Ten eggs with well-developed embryos that I examined, iden-
tified as B. bathymaster by E. H. Ahlstrom. collected in the
mouth of the Gulf of California^ ranged from 0.88-1 .00 mm in
diameter (.v = 0.94 mm) and had a single oil globule 0.22-0.28
mm in diameter (.v = 0.24 mm). The chorion was smooth, per-
ivitelline space narrow and yolk homogeneous. The oil globule
was situated in the posterior part of the yolk mass. Several small
melanophores were scattered on the head and dorsal side of the
embryo and on the ventral side of the tail.
Larvae — The larvae are not unusual. Their general morphology
is similar to that of other gadiform larvae but bregmacerotids
are not likely to be confused with them or with larvae of other
tropical-subtropical fishes with which they occur. In bregma-
cerotids, metamorphosis is gradual and direct.
Newly-hatched larvae are small, approximately 1.5 mm NL,
a fact often not appreciated when collecting nets with >333-
^m meshes have been used. The smallest larvae usually have
Table 83. Geographic Distribution Information and Some Meris-
Tic Data of Bregmaceros Adults and Larvae > 8 mm SL. Numbers
in parentheses are the most common counts for a species. For additional
meristic data, see Fahay and Markle, this volume.
' Ahlstrom, E. H. Personal Notes. "Gadiformes." Notes on file at
National Marine Fisheries Service, Southwest Fisheries Center, La Jolla,
California, USA.
' The eggs, identified as B. bathymasler. were provided by Dr. H. G.
Moser, Southwest Fisheries Center, National Mannc Fisheries Service.
La Jolla, California. They were collected on 10 June 1957, Station
I45G.40, Cruise 5706-S, near the mouth of the Gulf of California. I
could not confirm that the eggs were those of Bregmaceros. although
embryo myomere numbers were in the reported range for B. bathy-
master.
Myomeres
Dorsal
Anal
Species
Distnbution'
(vertebrae)
fin rays
fin rays
B. macclellandi
CTO
52-59
57-65
58-69
(54-55)
(58-61)
(62-66)
B. atlanticus
CTO(N)
50-55
47-56
49-58
(52-53)
(50-54)
(52-55)
(B. japonicusl)
WPO
56-58
51-60
56-63
B. nectabanus
I, IP, WPN
47-52
42-55
(47-50)
43-55
(50-52)
B. cantori
WAN
45-48
45-48
45-49
B. rarisquamosus
I. IP. WPN
43-48
(43-46)
34-41
36-43
B bathymaster
EPN
48-51
44-50
45-52
B arabicus
I. IP, WPN
50-54
50-60
(52-54)
50-63
(56-57)
B. Type A
WAN
44-47
40-44
42-46
' CT = circumlropical; O = oceanic; N = nentic; WP = weslem Pacific; I = Indian; IP = Indo-
Pacific; WA = weslem Atlanlic; EP = eastern Pacific.
not been described, although it is dunng that stage when specific
pigmentation is unique and identification easiest. Small speci-
mens (1.5-3.1 mm NL) of eight species are illustrated (Figs. 153
and 154). Larvae of 3.0-6.0 mm SL may be most difficult to
identify because pigment patterns are in transition and fin rays
have not developed sufliciently to be diagnostic. At lengths > 6.0
mm identification becomes easier, based on pigmentation char-
acteristics (Figs. 1 55 and I 56) and on complete (or nearly com-
plete) counts of median fin rays and myomeres. For larvae >10-
1 1 mm, diagnostic meristics usually are complete and illustra-
tions/descriptions in D'Ancona and Cavinato (1965) and Be-
lyanina (1974) usually will lead to correct identifications. Use
of information on larval pigmentation, meristics and size at
occipital ray development allow all described species to be iden-
tified.
Occipital ray (Table 84).— The size at appearance of the single
occipital ray varies among species. It is the first fin ray to develop
in B. macclellandi and in B. Type A (Houde, 1981), appearing
when larvae are 2.0-2.5 mm in length. In all other species the
ray develops at lengths of 5.0-7.5 mm, usually at approximately
6.5 mm. The occipital ray ofB. macclellandi is long and delicate,
often extending to near the middle of the second dorsal fin in
specimens < 1 0 mm, but subsequently declining in relative length.
In other species, the occipital ray is shorter, never reaching the
second dorsal fin.
Pigiyientation (Table 84) —'Larvae of the oceanic species B.
macclellandi and B. atlanticus are darkly pigmented. Larvae of
the neritic species are lightly pigmented. All have heavy internal
pigment over the visceral mass. The most distinctive pigment
is present on the smallest larvae (Figs. 153 and 154) and all
described species can be identified using pigmentation patterns
for larvae of 1.5-3.0 mm SL. The amounts of pigment, and
particularly the diagnostic melanophore patterns, tend to be lost
or reduced as larvae grow. External pigment tends to migrate
internally with growth, the tendency being most apparent in the
neritic Indo-Pacific species B. nectabanus. B. arabicus and B.
rarisquamosus. At the smallest lengths, the closely related B.
nectabanus and B. cantori have obviously different pigmenta-
Fm 1 53 Urvae of Bregmaceros in the length range 2. 1 to 3. 1 mm NL. (A) B macMland,. 3.0 mm. 2701 5'N, 084»28'W; (B) B allanUcus.
2.9 mm 27W 084»2Tw (C) B. Type A. 3.1 mm. 26°00'N. 083«53'W; (D) B. balhymasWr. 2.1 mm, 22°55'N, 108-40'W.
HOUDE: BREGMACEROTIDAE
303
Fig. 154. Larvae of Bregmaceros in the length range 2.1 to 3.1 mm NL. (A) B. nectabanus. 2.8 mm. 26°18'N. 052°00'E; (B) B. canlori, 2.6
mm; 27°00'N. 084°2rW; (C) B. arabicus. 2.5 mm, 29°26'N, 048°00'E; (D) B. rarisquamosus. 2.5 mm, 25°52'N, OSS'SS'E.
B
Fig 1 55 Larvae of Bregmaceros in the length range 7.0 to 10.0 mm SL. (A) B macclellandi. 7.0 mm, 1 3°00'N, 060°00W, (B) B. attanucus.
9.0 mm, 24''34'N, 082°56'W; (C) B. Type A, 8.5 mm, 27"'00'N, 084°22'W; (D) B. balhymaster. 9.5 mm. 13°I2'N. 09r5rw.
HOUDE: BREGMACEROTIDAE
305
Table 84. Size at Appearance of Occipital Ray and Pigmentation Characteristics of Bregmaceros Larvae in Two Length Ranges.
See Figures 153-156. In addition to pigment described, all Bregmaceros larvae have internal pigment on dorsal surface of visceral mass.
Species
Size at
appearance of
occipital ray
(mm SL)
^ .1 mm SL
Distinctive pigmentatii
5-10 mm SL
B. macclctlandi
B. atlanticus
B. nectabanus
B. canlori
B. ransquamosus
B bathyniastcr
B. arabicus
B. Type A
2.0-2.5 Melanophore at angle of jaw and tip of lower jaw;
scattered melanophores on head and at base of
pectoral fin. A few large, internal stellate mela-
nophores in double row on side of body and
tail. Melanophores on ventral surface of viscer-
al mass.
-5.0-5.5 Melanophore at angle of jaw and tip of lower jaw.
Scattered melanophores, on head and over
midbrain and at base of pectoral fin. Scattered,
large internal stellate melanophores on side of
body. Diffuse melanophores, some dendritic,
on surface of trunk and tail.
6.0-7.0 Single melanophores at angle of jaw, over hind-
brain, on nape and just anterior to anus. Dif-
fuse melanophores in short, double row on side
of tail and also in dorsal and ventral finfolds
directly above and below the double row. Me-
lanophore on ventral side of tail, just anterior
to notochord tip.
6.0-7.0 Melanophore at angle of jaw. A few small mela-
nophores on ventral surface of visceral mass.
6.0-7.0 Melanophore at angle of jaw and on lower jaw
tip; also over hindbrain. Few scattered melano-
phores on ventral surface of visceral mass. Dif-
fuse melanophores in three patches on side of
tail. Melanophore on dorsal surface of tail just
antenor to notochord lip.
6.5-7.5 Melanophore at angle of jaw and at anus. A few
scattered melanophores on ventral surface of
visceral mass. A row of 5-7 melanophores on
the ventral side of the tail.
6.0-7.0 Melanophore at tip of lower jaw and on ramus of
lower jaw (elongate melanophore). Often a few
scattered melanophores on ventral surface of
visceral mass. Diffuse pigment in three patches
on side of tail. Melanophore on side of tail just
anterior to notochord tip. Single melanophore
over forebrain.
2.0-2.5 Melanophore at angle of jaw and on tip of lower
jaw. Scattered melanophores over hindbrain.
Melanophores on occipital ray.
Small, scattered melanophores over surface of head
and body but not on posterior part of tail. Sev-
eral, large, internal stellate melanophores in a
double row on side of body and tail.
Many melanophores over surface of head and
body, including dorsal, anal and caudal fins. Lar-
va more or less "completely" pigmented.
Melanophore at angle of jaw. A few melanophores
on tail just anterior to its lip and sometimes one
or two melanophores at base of caudal fin. Inter-
nal melanophores on side of body, between
origins of dorsal and anal fins and also in tail
midway between origins of those fins and lip of
tail.
Melanophore at angle of jaw. A large melanophore
often present over forebrain. Internal pigment
visible near otoliths and just antenor to insertion
of pectoral fins. Internal pigment sometimes visi-
ble along developing vertebral column.
Melanophore at angle of jaw. Large, intense group
of melanophores in caudal fin. Scattered melano-
phores on anterior, ventral surface of visceral
mass. Some internal melanophores along devel-
oping vertebral column in tail, just antenor to its
lip.
Melanophore at angle of jaw and on ramus of lower
jaw. Melanophores on snout and on surface over
fore- and midbrain. Melanophore at anus and
two or more melanophores on dorsum just under
base of anterior third of second dorsal fin. In-
tense group of melanophores in caudal fin. Inter-
nal pigment along developing vertebral column
in peduncle region.
Melanophore at tip of lower jaw and an elongate
melanophore on ramus of lower jaw. Several me-
lanophores in base of caudal fin and a few scat-
tered melanophores on tail just anterior to caudal
fin.
Melanophore at angle of jaw. Scattered melano-
phores over fore-, mid- and hindbrain and on
nape. Melanophores on occipital ray and in pel-
vic fins. Scattered melanophores on ventral sur-
face of anterior half of visceral mass sometimes
present.
tion. The smallest B. macclellandi and B. atlanticus larvae po-
tentially could be confused, based on pigmentation alone. Breg-
maceros macclellandi has less external pigment and internal
pigment on tail and body is more clearly organized into two
rows than that of B. atlanticus.
As larvae grow pigmentation becomes less reliable as a means
to identify them. Nevertheless, the patterns are distinctive enough
to allow tentative identification (Figs. 1 55 and 1 56), which can
be confirmed by considering meristic characters. Because of
identification errors, there are erroneous descriptions of pig-
mentation in the literature. For example, larvae of fi. nectabanus
< 10 mm do have a melanophore at the jaw angle, although the
review literature indicates that it is absent (D'Ancona and Cav-
inato, 1965; Belyanina, 1974).
Meristic characters (Table 83). — Excepl for B. arabicus. neritic
species have lower myomere, vertebrae and median fin ray counts
than do B. macclellandi or B. atlanticus. The lowest meristics
occur in B. rarisquamosus and the highest in B. macclellandi.
The neritic B. arabicus has meristics similar to those of B.
Fig. 156. Larvae of Bregmaceros in the length range 7.0 to 10.0 mm SL. (A) B. nectabanus. 10.0 mm. 25''28'N, 053°50'E; (B) B. canton. 8.0
mm, 27°15'N, 083°53'W; (C) B. arabicus. 8.9 mm, 29°00'N. 048°29'E; <D) B. ramquamosus. 7.0 mm, 27°4rN, 049°45'E.
HOUDE: BREGMACEROTIDAE
307
atlanticus but larvae of the two species are easily separated by
pigmentation differences. There is slight overlap in meristics of
B. nectabanus and B. canton, although B. canton generally has
lower counts. The very wide range in dorsal and anal fin ray
counts attributed to B. nectabanus possibly has resulted from
identification errors.
Adult complements of median and of caudal fin rays are
present at 7.5-9.5 mm SL. Three or four pelvic fin rays develop
early in larvae, most precociously in B. macclellandi and B.
Type A, just after appearance of the occipital ray. In other
species pelvic rays appear at 3.5-4.5 mm length prior to ap-
pearance of the occipital ray. As larval development proceeds
an additional 2-3 pelvic rays ossify, giving the adult comple-
ment of 5-7 rays.
Relationships
Family relationships. — The bregmacerotids are gadiform fishes
(Fahay and Markle, this volume) of uncertain affinities and with
no obvious close relatives (Cohen, this volume), but generally
thought to be most closely related to the Muraenolepidae, Mor-
idae and Melanonidae (Nelson, 1976; Fahay and Markle, this
volume). Although affinities are unclear, bregmacerotids are
clearly gadiforms. They have high vertebral numbers (Table 83),
a long tail and long median fins with numerous rays (Cohen,
this volume; Fahay and Markle, this volume). A well-developed
caudal fin, separate from the dorsal and anal fins, is present.
Accessory (x and \) bones, believed to be a primitive character
in Gadiformes, are present in the caudal complex. But the num-
ber of hypurals has been reduced to two inferior elements and
a platelike superior element, believed to represent fusion of
hypural elements 3-5 (Markle, 1982; Cohen, this volume). The
caudal fin of bregmacerotids is the most symmetrical in the
Gadiformes. Both Ahlstrom" and Markle ( 1 982) have illustrated
the caudal skeleton of a Bregmaceros sp.; Markle's specimen is
undoubtedly B. macclellandi. based on meristics that he gives.
The number of principal (branched) caudal rays is 12, among
the lowest in gadiform fishes. Procurrent (unbranched) rays are
numerous, 20-24 in number, equally divided between the dorsal
and ventral sides of the caudal complex. One principal ray is
associated with each inferior hypural, 8 are associated with the
superior hypural plate and one is associated with each epural
bone. No uroneural is illustrated by Ahlstrom' but Markle ( 1 982)
illustrated one and noted that its presence is unique among
gadoid fishes. Six vertebral centra appear to be involved in
caudal fin ray support. The first dorsal fin, which consists of a
single, elongate ray, is located on the occiput, a unique condition
in gadiform fishes. The pelagic, tropical-subtropical distribution
of bregmacerotids is unusual among gadiforms.
Species relationships. — The species of Bregmaceros are remark-
ably similar. They have wide geographic distributions with little
apparent tendency to differentiate over their ranges of occur-
rence. Belyanina (1974) discussed the evolution and dispersal
of Bregmaceros. She believed that the family originated in the
Indo-Malayan Archipelago from which it dispersed with little
morphological modification. The present-day richness of species
in the Archipelago and the adjacent northern Indian Ocean lends
credence to that hypothesis. Five species (B. macclellandi, B.
atlanticus, B. nectabanus, B. rarisquamosus and B. arabicus)
presently occur in the proposed area of origin. Three species,
{B. bathymaster, B. contort and B. Type A) do not occur there.
The first two of these resemble B. nectabanus and may be de-
rived from it. The western Atlantic B. Type A is enigmatic
Bregmaceros
Proposed Species Relationships
B. rtectabanus
.B can fori
.B. bathymaster
B arabicus
B. rarisquamosus
B. macclellandi
B. atlanticus
B. japonlcus
'<>^\ /Bregmacerotidae
Fig. 157. Proposed species relationships of the Bregmacerotidae.
The possible relationships, indicated by the branching points, are based
on interpretations of species distributions and on meristic characters
and larval pigmentation.
because it differs substantially from all described species. Be-
lyanina (1974, 1980) believed that B. nectabanus was the com-
mon neritic Bregmaceros in the western Atlantic but subsequent
research (Milliken, 1975; Houde, 1981; Milliken and Houde,
1984) has demonstrated that the western Atlantic species, B.
cantori, differs substantially in modal vertebral numbers and
median fin ray counts, and also that the larvae differ significantly
in pigmentation characteristics.
Based on the species characteristics and known distributions,
possible relationships among species are proposed in Fig. 157.
Belyanina ( 1974) believed that the two oceanic species, B. mac-
clellandi and B. atlanticus, evolved from neritic species. It seems
equally probable that the neritic species evolved from the two
circumtropical, oceanic species. B. macclellandi and B. atlan-
ticus are very similar. They have relatively high meristic counts
and are darkly pigmented. Their larvae are heavily pigmented
and tend to be deeper-bodied than larvae of neritic species. The
neritic species, except B. arabicus. have vertebral numbers and
median fin ray counts much lower than those of B. macclellandi
and B. atlanticus. As larvae the neritic species are relatively
thin-bodied and lightly pigmented (Table 84, Figs. 1 5 3 and 1 54).
Bregmaceros nectabanus. B. arabicus and B. rarisquamosus
overlap broadly in their ranges of occurrence, as do B. mac-
clellandi and B. atlanticus and, to a lesser extent, B. cantori and
B. Type A. Species frequently are collected together as larvae
308
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
in ichthyoplankton surveys. Only B. hathymaster appears to
live in relative isolation from other species of Bregmaceros. In
the Indo-Pacific, B. ransquamosus (small-size, early matura-
tion, low meristics) and B. arabkus (high meristics) possibly
were derived from a B. nectabanus stock intermediate in me-
ristic characteristics. The basic B. nectabanus stock also may
have given rise to B. bathymaster and B. cantori. Detailed study
of eastern Atlantic B. cantori-\'\V.e larvae may help to resolve
questions about dispersal and evolution of species.
Bregmaceros Type A is curious. Like B. macclellandi. its lar-
vae develop the occipital ray at <2.5 mm (Table 84). Yet, it
bears little resemblance to B. macclellandi in other meristic or
pigmentation characters. It has the lowest vertebral and median
fin ray counts of any Bregmaceros except B. ransquamosus (Ta-
ble 83). Larvae of Type A generally occur over the deep shelf
and slope, occasionally in oceanic waters, and often co-occur
with B. cantori and B. atlanticus (Houde, 1981).
The status of B. japonicus is unclear although this form may
be a western Pacific variety or subspecies of B. atlanticus (Be-
lyanina, 1974). A recent reexamination of the holotype (Masuda
and Ozawa, 1979) indicated that its vertebral and median fin
ray counts exceeded or were at the upper extreme of ranges
reported for B. atlanticus (Table 83). There is a need for critical
examination of B. atlanticus and B. japonicus specimens from
the tropical Pacific Ocean. Juveniles and adults that I examined'
from the eastern, tropical Pacific appeared to be typical B. at-
lanticus but none of the small larvae had typical B. atlanticus
pigmentation. A moderately heavily-pigmented larva was pres-
ent in tropical Pacific collections that may be an undescribed
species. Its status and its possible relationship to the B. atlan-
ticus/B. japonicus systematics problem need to be determined.
University of Maryland, Center for Environmental and
EsTUARiNE Studies, Chesapeake Biological Laboratory,
Solomons, Maryland 20688.
Ophidiiformes: Development and Relationships
D. J. Gordon, D. F. Markxe and J. E. Olney
THE order Ophidiiformes contains 300-400 species occu-
pying mostly benthic habitats over a broad range of depth
and salinity. These are elongate, tapering fishes with or without
a caudal fin. The dorsal and anal fins are long, sometimes con-
fluent, without spines and with pterygiophores more numerous
than adjacent vertebrae. The pelvic fins, if present, are located
far forward and are reduced to one or two rays, sometimes with
a small spine.
Cohen and Nielsen (1978) summarized the present under-
standing of the systematics of ophidiiform fishes, presented keys
to the genera, and provided a useful framework on which to
base a discussion of the order. The presence or absence of vi-
viparity defines two suborders, Bythitoidei and Ophidioidei.
Bythitoidei contains the live-bearing "brotulids" and is divided
into two families, Aphyonidae and Bythitidae. The oviparous
Ophidioidei contains Ophidiidae and Carapidae. Ophidiidae
includes the cusk-eels (Ophidiinae) and the oviparous "brotu-
lids," previously allied with the bythitoids in the family Bro-
tulidae.
Aphyonidae, reviewed by Nielsen (1969), contains 18 species
in five genera. These ovoviviparous fishes are benthopelagic and
found worldwide. Bythitidae contains over 80 species in 28
genera. Most species of this family occur either in shallow trop-
ical waters, including coral reefs, or in waters of intermediate
depths on the continental shelf and slope. Some deeper-dwelling
slope species occur at higher latitudes, a few species inhabit
abyssal waters and some are found in freshwater. Carapidae
contains about 30 species divided into two subfamilies (Pyra-
modontinae, Carapinae) and six genera, all possessing a vexil-
lifer larva (Olney and Markle, 1979; Markle and Olney, 1980;
Markle et al., 1983). Some species are free-living while others
are inquilines within the body cavities of invertebrate hosts
(Trott, 1 970; Trott, 1981). Ophidiidae, as defined by Cohen and
Nielsen (1978), includes oviparous ophidiiform fishes lacking a
vexillifer larva and possessing a supramaxillary bone. The fam-
ily is divided into four subfamilies: Brotulinae, Brotulotaeni-
inae, Neobythitinae and Ophidiinae. Brotulinae contains one
genus (Brotula) with at least five species (Cohen and Nielsen,
1978). Adult Brotula are benthic and circumtropical on the
continental shelf Brotulotaeniinae contains the single genus
Brotulotaenia with four midwater, tropical representatives (Co-
hen, 1974). Neobythitinae is a morphologically diverse group
containing 38 genera and over 135 species with worldwide dis-
tribution and a wide depth range, but mostly deep sea. Ophi-
diinae consists of about 60 nominal species with several un-
described forms (Lea, 1980), mostly in shelf waters.
Development
Knowledge of the early life history of ophidiiform fishes varies
considerably among major taxa. Larvae of the live-bearing species
are infrequently collected and larvae of deep water forms are
even rarer. The incomplete state of knowledge of the taxonomy
of bythitoid fishes renders identification of most of their larvae
tentative. On the other hand, carapid and ophidiine larvae are
common to abundant in tropical plankton. Carapid larvae are
relatively well known and have proven to be of systematic value
(Olney and Markle, 1979; Markle and Olney, 1980). Though
the larvae of only a small percentage of the species of ophidiines
are known, these larvae provide useful characters for under-
standing relationships within the group (Gordon, 1982).
GORDON ET AL.: OPHIDIIFORMES
309
Fig. 158. Larvae of Bythitoidei. (A) Larva of Brosmophyas marginala. 12.5 mm NL. NMFS-SWFC. CalCOFI 7207 .A.x Sta. 63.52. (B)
Unidentified bythitid larva, 21.9 mm SL, HML H 4086, 40°34'N, 66''00'W. (C) Exterilium larva tentatively assigned to Neobythitinae, 29.5 mm
SL, MCZ-WHOl, Oceanus 22, JEC 771 1, 0°00'N, 37''40'W.
Eggs and embryos— Ophidiiform eggs are poorly known. The
pelagic eggs of Gcnypterus capensis (Ophidiidae) are moderately
large, spherical and contain a single oil globule (Brownell, 1 979).
The few known carapid eggs are pelagic, ellipsoidal, and possess
a single oil globule. Early developmental stages may be con-
tained in a mucilaginous raft. Eggs have been described for
Carapus acus {Emtry, 1880; Padoa, 1956j). Echtodon dentatus
(Sparta, 1926), E. c/n/wwow^/ (Kennedy and Champ, 1971), E.
rendahli (Robertson, 1975b), and unidentified carapid species
from the North Atlantic (Ryder. 1884) and South Africa (Brow-
nell, 1979).
Aphyonid larvae have not been reported from plankton tows
but late embryos taken from ovarian tissue were illustrated by
Nielsen (1969).
310
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
5sSSS««*§^'*5''^^^^^^^5^5S'?S?5S^^
S*^!«S^!5SS!«Ssi5aS5
■"■■■ --^^^<^^^^^^^^^^^^<^^^^^^^ -,s=,^,s.*j.^*^-^'
GORDON ET AL.: OPHIDIIFORMES
311
D
Fig. 1 60. Larvae of the genus Lepophidiuln. (A) Lepophidium negropinna. 5. 1 mm NL. NMFS-SWFC, P-28 1 20.35. (B) Lepophidiumjeannae,
1 1 mm SL. UMML CI 71 13 Sla. 62, 26°30'N. 83°00'W. (C) Lepophidium staurophor. 12 mm SL. UMML, CI 71 13 Sta. 81, 27°00'N, 84''05'W.
(D) Lepophidium Type 1, 7.8 mm NL, UMML, CI 71 14 Sta. 127, 28''15'N, 84°50'W.
Larvae. — The reproductive biology of three bythitid species has
been discussed (Wourms and Bayne, 1973; Wourms and Cohen,
1975; Suarez, 1975). Aboussouan (1972a) described a larva at-
tributable to Oligopus longhursti and Leis and Rennis (1983)
have illustrated a larval Dinematichlhys. A larva of Brosino-
phycis marginata from the eastern Pacific and an unidentified
bythitid from the North Atlantic (Fig. 1 58) are illustrated here.
Larvae of a number of carapid and ophidiine species have been
described, but few larvae of other, generally deeper-dwelling,
ophidioid taxa are known. Leis and Rennis (1983) illustrated a
larval Brolula. Aboussouan (1980) described a large, ribbon-
shaped larva which he attributed to Bwtidotaenia. A specimen
of Spectrunctulus grandts (56 mm SL) is illustrated and dis-
cussed by Nielsen and Hureau (1980). Larvae of the neobythi-
Fig 159. Urvae of tribe Ophidimi. (A) Olophidium omosligmum. 8.3 mm NL, UMML, CI 71 14 Sta. 126, 28°15'N, 84°25'W. (B) Ophidian
Type 1, 7.6 mm NL, UMML, CI 7308 Sta. 60, 26°3rN, 82°28'W. (C) OphidionType 2, 7.0 mm NL, UMML, CI 7303 Sta. 94, 27°30'N, 83°29'W.
(D) Ophidian nocomis. 24 mm SL, NMFS-SEFC, Ore II 7343 Sta. 160. (E) Ophidian setenops. 24 mm SL, UMML, CI 71 13 Sta. 95, 27''31'N,
83°46W. (F) Parophidion schmidli. 17 mm SL, MCZ-WHOI, RV Chain 60 RHB 1315, 25°46'N, 79°47'W.
312 ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 85. Meristic Variation in Western North Atlantic Species of Cusk-Eels. Sample size is indicated in parentheses below the range.
Vertebrae
Fin i^ys
Species
Precaudal
Caudal
Total
Dorsal
Anal
Pectoral
Source
Otophidium omostigmum
14
42-44
56-58
99-108
84-87
16-18
Gordon. 1982; Bohlke
(10)
(10)
(10)
(10)
(10)
(20)
and Robins. 1959
Otophidiuin dormilator
14, 15
49-51
64,65
111-117
94-96
15-16
Gordon. 1982; Bohlke
(11)
(7)
(10)
(14)
(14)
(20)
and Robins. 1959
Otophidium chickcharney
13
50-52
63-65
111-116
98-102
16, 17
Bohlke and Robins.
(6)
(6)
(6)
(9)
(8)
(16)
1959; onginal
Parophidion schmidti
15
52
67
115-126
98-106
17-19
Bohlke and Robins.
(1)
(1)
(1)
(24)
(24)
(25)
1959; original
Ophidian selenops
15, 16
62-65
77-81
132-138
123-129
15-16
Gordon. 1982; Robins
(14)
(14)
(14)
(8)
(8)
(18)
and Bohlke, 1959
O. nocomis
17
67-70
84-87
144-153
132-139
14-17
Robins and Bohlke,
(7)
(7)
(7)
(42)
(42)
(74)
1959
O. holbrooki
15, 16
50-53
66-69
117-132
97-109
19-21
Gordon. 1982
(12)
(12)
(12)
(12)
(12)
O. beani
15-17
50-53
65-69
111-133
94-103
18-21
Gordon. 1982
(45)
(45)
(45)
(9)
(9)
O. grayi
16
48-50
64-66
131-145
99-113
20-23
Gordon. 1982
(11)
(11)
(11)
(11)
(11)
O. marginatum
15
53-54
68-69
147-158
118-124
—
Miller and
(4)
(4)
(4)
(4)
(4)
Jorgenson, 1973
O. welshi
15, 16
50-52
67,68
128-150
105-122
21
Gordon, 1982
(6)
(6)
(6)
(15)
(15)
O. lagochila
—
—
—
123-125
103-105
17-19
Bohlke and Robins,
(3)
(3)
(6)
1959
Lepophidium graellsi
14-16
55-57
69-73
124-133
101-109
20-22
Gordon. 1982
(40)
(39)
(32)
(38)
(37)
L. marmoratum
14, 15
55-60
70-75
121-136
103-112
21-24
Gordon, 1982
(31)
(31)
(32)
(30)
(31)
L. jeannae
14, 15
58-60
73-75
131-140
112-117
20-21
Gordon, 1982;
(9)
(9)
(9)
(11)
(11)
(12)
Robins, 1960
L. staurophor
15
65-67
80-82
140-147
122-127
22,23
Robins. 1958
(3)
(3)
(3)
(4)
(4)
(4)
L. profundorum
15-17
58-61
73-78
131-140
110-121
22-24
Gordon. 1982
(14)
(14)
(14)
(14)
(14)
L. kallion
15
59
74
133
108
23,24
Robins, 1959
(1)
(1)
(1)
(1)
(1)
(4)
L. aporrhox
13
52.53
65.66
109-114
96-99
21-23
Robins, 1961
(5)
(5)
(5)
(7)
(7)
(14)
L. pheromystax
14. 15
54-57
69-72
125-132
104-110
20-22
Robins, 1960
(20)
(20)
(20)
(41)
(41)
(86)
tine genus Benthocometes {Pteridum) were illustrated by Padoa
(1956i). Exterilium larvae (Fraser and Smith, 1974;Moser, 1981)
may be larvae of deep-dwelling neobythitine species (Figure
158C).
Larvae of six ophidiine genera are known. Padoa (19561)
described the larvae of Parophidion vassali and Ophidian bar-
baturn from the Mediterranean Sea. Aboussouan (1972a) de-
scribed the larvae of Ophidian barbatum from the eastern At-
lantic. Brownell (1979) reared early stages of the larvae oi Gen-
ypterus capensis. Larval stages of Ophidian marginatum were
illustrated in Scotten et al. (1973) and were reproduced in
Fritzsche (1978). Figure 178b in Fntzsche (1978) is probably a
stichaeid and not an ophidioid species. Larval stages of Oto-
phidium omostigmum (Fig. 1 59A), Ophidion selenops (Fig. 1 59E),
Lepophidium jeannae (Fig. 160B) and Lepophidium staurophor
(Fig. 160C) from the Gulf of Mexico were described by Gordon
(1982). Larvae of Chilara taylari and Ophidian scrippsae were
described by Ambrose et al. ( 1 983). Larvae of Ophidion nocomis
(Fig. 159D) and Lepophidium negropinna (Fig. 160A) are illus-
trated in this study.
Carapid larvae have been described (or illustrated) for six
genera and 12 species: Carapus acus (Psidoa. I956j); C. imberbis
(Aboussouan, 1972a); C. berniudensis {Dawson, 1971b; Olney
and Markle, 1979); Echiodon dentatus (Emery, 1880; Sparta,
1926; Padoa. 1947; Maul. 1976); E. drummondi (\ihrtinha\im,
1905-1909; Kennedy and Champ, 1971); Echiodon rendahli
(Robertson, 1975b); E. dawsoni (Olney and Markle, 1979); E.
c.xsilium (Trott, 1970; Olney and Markle, 1979); Encheliophis
jordani (Tron, 1970); Onuxodon margaritiferae (Gowoni et al.,
1984); Snyderidia canina (Strasburg, 1965; Markle and Olney,
1980); and Pyramodon wnlralis (Markle and Olney, 1980). In
some cases, larval identifications are unsubstantiated and cau-
tion should be employed, especially in the older literature. For
example, Padoa ( 1 956j) confuses larval E. dentatus (plate XLIV,
GORDON ET AL.: OPHIDIIFORMES
313
Table 86. Meristic Variation in Eastern North Pacific Species of Cusk-Eels. Sample size is indicated in parentheses below the range.
Vertebrae
Fin rays
Species
Precaudal
Caudal
Toial
Dorsal
Anal
Pectoral
Source
Otophidium indefatigabile
13-15
45-49
59-64
(17)
106-115
(15)
88-96
(14)
18-19
Lea,
1980
Chilara taylori
18-19
68-72
86-91
(66)
187-229
(50)
150-181
(50)
24-26
Lea,
1980
Ophidian costaricense
14-16
50-54
65-69
(45)
130-153
(44)
107-128
(43)
23-26
Ua,
1980
O. fulvum
13-15
49-55
63-69
(39)
137-160
(34)
112-136
(34)
23-26
Lea,
1980
O. galeoides
14-17
47-49
61-64
(52)
123-143
(42)
97-114
(41)
21-23
Lea,
1980
O. imitator
15-16
55-60
70-76
(20)
135-163
(18)
112-139
(18)
25-28
Lea,
1980
O. ins
16-17
53-56
69-73
(67)
121-148
(65)
98-122
(63)
22-24
Lea,
1980
O. moche
15-16
55-58
71-74
(6)
142-148
(6)
118-126
(6)
24-25
Lea,
1980
O. scrippsae
14-16
50-54
65-69
(102)
124-153
(102)
99-126
(100)
20-23
Ua,
1980
Lepophidium prorates
14-16
55-58
70-73
124-133
106-113
21-24
Robins, 1962
(63)
(59)
(59)
(59)
(60)
(87)
L. pardale
15
57
71
128, 132
106, 109
22,23
Robins, 1962
(2)
(2)
(2)
(2)
(2)
(2)
L. stigmatislium
14. 15
55,56
70
125, 130
103, 109
19-21
Robins, 1962
(2)
(2)
(2)
(2)
(2)
(4)
L. microlepis
14-16
51-62
66-77
117-141
97-121
21-26
Robins and Lea, 1978
(83)
(83)
(83)
(81)
(82)
(95)
Cherublemma emmelas
13, 14
42-44
55-58
99-113
81-93
24-26
Robins, 1961
(14)
(14)
(14)
(33)
(32)
(58)
L. negropinna
15, 16
59-61
75,77
138-148
116-121
21-24
Robins, 1962
(5)
(5)
(5)
(12)
(12)
(22)
Fig. 14; plate XLV, Figs. 3 and 4) with C. acus (plate XLV, Fig.
5).
Prejuveniles. —Ophidiine lai^ae are pelagic, and the develop-
ment of most species proceeds directly without an abrupt tran-
sition period. The larvae of Ophidian selenops, however, are
extremely elongate and attain a length of 40 mm SL before
reduction of intervertebral spaces causes a reduction in total
length to about 24 mm (Gordon, 1982). Soon after this trans-
formation the body shape approaches the juvenile form and the
larvae become benthic. Elongate larvae identified here as O.
nocomis have a similar morphology (Fig. 1 59D). Most ophidiine
species probably become benthic at about 25-30 mm SL. Chi-
lara taylori and Farophidion, however, have extended nektonic
prejuvenile stages (Lea, 1980). The prejuvenile stage of C tay-
lori reaches 80 mm SL and was described as Ophidion nova-
culum by Harry (1951).
A specialized prejuvenile stage, known as a tenuis larva, has
been described for some Carapus species and Echiodon dawsoni
and is characterized by the absence of a vexillum and an initial
lengthening and subsequent reduction in total length (Emery,
1880; Arnold, 1956; Padoa, 1956j; Strasburg, 1961; Hipeau-
Jacquotte. 1967;Gustato, 1976;Trott, 1981; Williams and Shipp,
1982). The stage is poorly known and has been reported as an
obligate inquiline parasite (Trott, 1981), a free-living benthic
form (Trott. 1981), and a pelagic form sometimes attracted to
nightlights (Smith et al., 1981).
Meristic characters. — MensUc characters that are observable in
ophidiine larvae include myomere number, vertebral formula
(precaudal plus caudal vertebrae), dorsal fin ray number and
anal fin ray number. Pectoral fin ray number and gill raker
development, which are important taxonomically in adults, can-
not be considered complete in pelagic larvae. Meristic characters
show large variation within species. In many cases, published
ranges for these characters are based upon too few specimens
to accurately depict the range of variation. In addition, meristic
data show broad range overlap between several species and
caution should be employed. Positive identification of larvae
based solely on meristic characters, however, can be made for
some western Atlantic species, including Ophidion selenops. O.
nocomis, Otophidium omostigmum. Oto. dormitator. Oto.
chickcharney and Lepophidium staurophor. In the eastern Pa-
cific, larvae of Chilara taylori and Otophidium indefatigabile
are identifiable based on meristics. Ranges of meristic characters
for western Atlantic ophidiines are given in Table 85 and for
eastern Pacific ophidiines in Table 86. Several species of the
genera Ophidion and Lepophidium from the western Atlantic
are presently undescribed. and taxonomic questions remain to
be resolved (C. R. Robins, pers. comm.).
Development of the ophidiine vertebral column and fins was
described by Gordon ( 1 982). Total myomeres in larvae compare
closely with total vertebrae in adults. The number of preanal
myomeres present prior to coiling of the gut is usually greater
than the number of precaudal vertebrae in adults, because the
314
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 87. Meristic Variation in Selected Species of Carapid Fishes. Abbreviations used are: N — number of specimens examined; D,o—
number of dorsal rays whose bases lie anterior to 31st vertebra; Ajo— number of anal rays whose bases he anterior to 31st vertebra; P, — pectoral
rays; P^— pel vie rays; PCV—precaudal vertebrae; NVD— number of vertebrae to dorsal origin; NVA— number of vertebrae to anal origin; ARDO—
number of anal rays to dorsal ongin; NA— not applicable.
Species N D^ A^ P^ p] PCV NVD NVA ARDO
Pvramodon ventralis
10
48-52
46-53
27-29
1
15-18
5-8
6-8
NA
Snydehdia canina
4
49-51
44-46
25
abs
14-15
6-7
9-10
NA
Carapus acus
2
37-39
58
abs
18
11-12
3-4
20-21
Carapus inourlani
3
38
56-57
18
abs
15-16
12-13
2-3
22
Carapus homei
2
33-37
56-60
17-19
abs
16-17
14
3-4
24-26
Carapus pan^ipinnis
4
35-38
50-53
15-19
abs
16
12-13
4-5
17-19
Echiodon drummondi
6
42-45
47-49
15-17
abs
25-29
8-9
6-8
5-7
Echiodon cryomargarites
24
37-40
46-50
19-21
abs
25-29
11-12
6-8
9-12
Echiodon dawsoni
1
28
39
17-18
abs
22
12
6
12
Onuxodon parvibrachium
4
44-46
44-48
14-15
abs
16-17
6-7
6-8
1-2
Onuxodon margantiferae
6
46-47
46-50
16
abs
19-22
6-8
5-6
3-4
Encheliophis vermicularis
3
25-26
39-41
abs
abs
21-22
16-18
4-5
18-21
Encheliophis gracilis
4
28-31
45-50
15-17
abs
26-31
16-17
3-7
23-26
gut migrates forward by 2-4 myomeres during formation of the
gut coil (Gordon, 1982). By 8-10 mm NL, the haemal arches
are closed allowing accurate determination of the vertebral for-
mula in cleared and stained larvae. Rays of the dorsal and anal
fins develop from posterior to anterior. Development begins at
7-10 mm NL and is complete by 15-20 mm SL. The adult
complement of nine caudal rays and seven branchiostegal rays
is present by 10 mm NL in most species. These structures do
not appear until 1 5 to 20 mm SL in the elongate O. selenops.
The number of pectoral fin rays ranges from 16-28, with sizes
at which the first rays appear ranging from 1 3-20 mm SL. The
pectoral fin is complete in some species by 18-20 mm SL.
Traditionally, meristic characters have not been widely used
in adult descriptions of carapids (Arnold, 1956) and conse-
quently, some easily observed and useful characters such as
pectoral fin ray counts (Cohen and Nielsen, 1978; OIney and
Markle, 1979) can not be obtained from the literature. Inno-
vative meristics, partly borrowed from eel systematics (Nielsen
and Smith, 1978), are useful aids to identification of larval and
adult carapids (OIney and Markle, unpublished data). Table 87
summarizes some of these meristic characters for selected species
of carapids.
Morphology. — Except for the larvae of O. selenops and O. no-
comis, known ophidiine larvae show little variation in size,
shape and development (Figs. 159, 160). The larvae, which
hatch at 2-3 mm NL, are moderately elongate and taper slightly
from the head to the end of the notochord. The eyes are round
and conspicuous; the mouth is oblique. Larvae become more
laterally compressed with growth. In all species examined by
Gordon (1982), the gut is straight at hatching and develops a
single coil at 5-7 mm NL as a downward loop twists, displacing
the more posterior portion of the gut to the right.
Carapid vexillifers (Figs. 161-163) are elongate larvae with a
moderately sized head, large eye and nasal rosette, coiled gut,
short preanal length, and tapered body frequently ending in a
broken filament, and an elongate larval dorsal fin ray (vexillum)
in front of the first adult dorsal fin ray (OIney and Markle, 1979;
Govoni et al., 1 984). Larvae of Pyramodon and Snyderidia (Fig.
1 63) have a somewhat deeper head and trunk, shorter pre-dorsal
distance, relatively long anal fin pterygiophores, and more pec-
toral fin rays than other carapid larvae (Markle and OIney,
1980). Variations in gross morphology in carapine larvae are
limited to variation in gut shape and fin ray or vexillum position
(Figs. 161, 162; Table 87).
Pigmentation. — VxgmenXdiXion of ophidiine larvae is useful for
identifying species and species groups though care must be taken
since ontogenetic changes occur (Gordon, 1 982). Head pigmen-
tation typically consists of two or three melanophores present
distally on the suspensorium near the articulation with the lower
jaw. Abdominal pigmentation is usually variable within species
and consists of melanophores scattered ventral to the gut. Pig-
mentation on the posterior half of the body is the most useful
for taxonomic purposes. All Lepophidium larvae have 2-10
large spots placed medially along the base of the anal finfold
and 1-2 spots dorsally in the caudal peduncle region. Unlike
Lepophidium, larvae of Ophidion, Otophidium and Chilara have
patterns of small stellate melanophores present laterally on the
body. Several species of these genera can be recognized on the
basis of postanal pigmentation. Some species have larvae which
are very similarly pigmented: O. selenops and O. nocomis; O.
welshi and O. marginatum. Larvae of other species cannot be
distinguished using pigmentation or meristic characters (Gor-
don, 1982): Ophidion Type 1 (Fig. 159B) which represents O.
holbrooki, O. beani and unidentified Ophidion species; Ophidion
Type 2 (Fig. 1 59C) O. welshi and O. grayi; and Lepophidium
Type 1 (Fig. 160D) L. graellsi and L. jeannae.
Carapid vexillifers are sparsely to moderately pigmented (Figs.
161-163). Red chromatophores have been noted in fresh ma-
Fig. 161 . (A) Urva of Echiodon drummondi, 76.5 mm TL, ZMUC uncat., DANA St. 8371, 5 r29'N, 1 2''50'W. (B) Urva of Echiodon rendahli.
72 mm TL, CSIRO uncat., Warreen Cruise, Sta. 266/39, 36°17'S, 150°25'E. (C) Dorsal fin and vexillar supports of Echiodon dawsoni. 4.2 mm
HL and 10 mm HL. Abbreviations used are vex — vexillum, fdp— first dorsal pterygiophore. (D) Dorsal fin and vexillar supports of Echiodon
drummondi. 4.6 mm HL and 24.3 mm HL. Abbreviations as in Fig. 16 IC.
GORDON ET AL.: OPHIDIIFORMES
315
42mm HL
fdp
46 mm HL
D
/
fdp
10 mm HL
24 3 mm HL
316
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 162. Larva of Carapus sp., (top) uncat., Anton Bruun, CH'S, 65°03'E, 27-28 May 1964. Larva of Onuxodon panihrachium. (bottom)
ZMUC uncat., Dana St. 3768 XVI, 400 mwo, 1°20'S, 138°42E, 0315 hrs. 25 July 1929.
GORDON ET AL.; OPHIDIIFORMES
317
■'^■'^^^--^■<'«l/4'0^///////.'^^
///y/^/^y^^y.
^■s-^--^.
\\\\\V^wvv\\\^^n;^v;x;n^\^\\xx^-^v.-xx-
•^y'^^r^-
Fig. 163. Larva of Pyramodon ventralis. (top) uncat., 21°20-30'N, 158°20-30W, 19 Dec. 1977. Larva oi Snydehdia bothrops. (bottom) MCZ
uncat., RHB 1263, Chain 60, 12°58'N, 73°34W, 29 May 1966, 0-120 m, IKMT.
terial of Echiodon dentatus (Padoa, 1956j) and E. dawsoni (Ol-
ney and Markle, 1979) but are not normally retained in pre-
served material. Meianophores are variously present at the
symphysis of the lower jaw, on the snout, head, vexillum, swim-
bladder, trunk, and tail. Preliminary studies (Padoa, 1 956j; Rob-
ertson, 1975b; OIney and Markle, 1979; Markle and Olney,
1980) indicate that pigmentation may be regionally useful as an
aid to identification but seems problematic as an indicator of
higher relationships.
Osteology.— The placement of the pelvic fins, which defines the
subfamily Ophidiinae, shows marked ontogenetic change (Gor-
don, 1982). In early larvae, the cleithra lack the forward exten-
sion and the pelvic fins (appearing by about 7 mm NL) are
supported in a jugular position. By 20 mm SL, the bony exten-
sion of the cleithra develops and begins to elongate anteriorly.
The pelvic fins, which are supported at the symphysis, migrate
forward and are present in the characteristic mental position in
the juveniles. The presence of pelvic fins in the jugular position
has occasionally caused the confusion of early larvae with other
ophidioids.
The general structure of the vexillar ray is described by Olney
and Markle (1979) and Govoni et al. ( 1 984). External variations
of vexilla are in length, ornamentation, pigmentation, and po-
sition. Some variation such as length and ornamentation ap-
pears to be an artifact (Govoni et al., 1984). In several species,
the vexillar pigmentation and ornamentation are curiously re-
peated in the caudal filament (Fig. 162). Variation in the sup-
porting proximal radial is seen in its shape, its position relative
to the first adult dorsal fin ray and to vertebrae, and in fusion
with the proximal radial of the first dorsal fin ray. In addition,
the supporting proximal radial may or may not be retained in
adults (Fig. 161C, D). Its retention provides a means of iden-
tifying the location of the vexillum and can aid in larval iden-
tification. Its absorption, however, appears to have occurred
independently in several genera. In the pyramodontines (Markle
and Olney, 1980) and Carapus (Olney and Markle, 1979) there
is also an accessory cartilage in front of the second neural spine.
Its origin and function are unknown. Carapus (and presumably
Encheliophis) and the pyramodontines also have the most for-
wardly placed vexilla, usually above or in front of the first anal
fin ray. Carapus (and presumably Encheliophis) differs from the
pyramodontines and all other carapids in displacement of the
first adult dorsal fin ray far posteriad of the vexillum (Fig. 162).
Modified ribs on the anterior vertebral centra of carapids and
ophidiines are associated with sound production/reception (Rose,
1961; Courtenay and McKittrick, 1970; Courtenay, 1971) and
develop in early stages in carapids (Olney and Markle, 1979).
In Carapidae, the first two ribs are movable and all subsequent
ribs are rigid (Markle et al., 1983). A simple recurved third rib
318
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
is found in Echiodon and pyramodontines while an expanded
third rib is found in Carapus (Olney and Markle, 1979; Markle
and Olney, 1980; Williams and Shipp, 1982) and Oimxodon
(Courtenay and McKittrick, 1 970). The sexually dimorphic and
interspecific differences in swimbladder morphology of ophi-
diines appear only in juveniles and adults and are not useful in
distinguishing larvae.
The visceral cradle, formed from the criss-crossing elongate
proximal pterygiophores of the anterior anal fin rays, is a unique
specialization of Pyramodon (Markle and Olney, 1980). Its pre-
sumed precursor, non-crossing elongate proximal pterygio-
phores, is found in larval Snydehdia. The elongate proximal
pterygiophores found in pyramodontines are conspicuous in
larvae.
The pectoral fin of carapids is a variable structure and po-
tentially useful in the study of relationships as well as for iden-
tification. Adults of some species of Encheliophis completely
lack a pectoral fin while pyramodontines have a well-developed
fin with up to 29 rays. Most cleared and stained carapid and
ophidiine larvae have an elongate, cartilaginous, ventral process
of the coracoid (VPC). In the carapid "exterilium" larvae (Fig.
16 IB, see also Robertson, 1975b) the development of the VPC
has been carried to an extreme. The hanging or trailing gut of
this larva is supported by a skeleton of the two VPC's which
intertwine with the intestine. Support of a trailing gut by VPC's
is not unique since we have also seen it in the ophidioid "ex-
terilium" (Fraser and Smith, 1974; Moser, 1981) and Symphii-
rus minor (unpublished data).
The dentition of carapids is useful for adult identification
purposes (Arnold, 1956) and enlarged canines as well as the
dentary diastema have been used to separate Carapus and
Echiodon larvae (Olney and Markle, 1979).
Relationships
Intra-ordinal relationships.— T\\t classification of Ophidi-
iformes proposed by Cohen and Nielsen (1978) differs most
significantly from earlier classifications in the use of mode of
reproduction as a subordinal character. Previous classifications
recognized the highly specialized carapids as either one or two
families (Carapidae and Pyramadontidae) and, based on the
position of the pelvic fins, divided the remaining ophidiiforms
into two groups, the ophidiids (ophidiines, pelvics mental) and
the brotulids (pelvics absent or jugular).
Relationships within the Bythitoidei remain unclear. The
aphyonids share a number of neotenic characters serving to
define the family. This may be a polyphyletic group, however,
with common character states reflecting convergent trends (Co-
hen and Nielsen, 1 978). Comparisons ofembryonic adaptations,
such as trophotaeniae (Wourms and Cohen, 1975), may prove
useful in resolving systematic problems within Bythitidae. Two
subfamilies (Bythitinae and Brosmophycinae) are presently de-
fined on the basis of confluence of anal and dorsal fins with the
caudal fin, though neither subfamily has been adequately stud-
ied.
Ophidioidei is defined by the presence of oviparity and the
anterior nostril (in most genera) well above the upper lip. The
relationships of the ophidioid subfamilies are also uncertain and
the suborder may be paraphyletic. Carapidae and subfamily
Ophidiinae each seem to form natural groupings based upon
well-defined synapomorphies. Further study of the neobythi-
tines may reveal several natural groupings (Cohen and Nielsen,
1978). The relationships of Brotulotaeniinae and Brotulinae are
unknown.
Two tribes of Ophidiinae can be defined on the basis of squa-
mation and the presence of pyloric caecae. Lepophidiini (im-
bricate scales; pyloric caecae present) contains three genera: the
monotypic Cherublemma emmelas. Genypierus, and Lepophi-
diiim. Lea (1980) has proposed the elevation of Genypterus to
the level of tribe. The Ophidiini (anguilloid squamation; pyloric
caeca absent) contains the genera Ophidian, Otophidium. Chi-
lara, Raneya and Parophidion. These genera, established on the
basis of meristics. morphometries, swimbladder morphology
and squamation, are not well-defined and require further study.
A comparative study of the development of ophidiine larvae
of three nominal genera, Ophidion. Otophidium and Lepophi-
dium, suggests that body shape, development of the caudal fin
and pigmentation can provide useful taxonomic characters
(Gordon, 1982). The body shape and development of Lepo-
phidium larvae may represent the primitive state for the subfam-
ily. Otophidium omostigmum and most Ophidion species retain
this morphology, as does Parophidion (Fig. 1 59F; Padoa, 1 956i).
The morphology and development of O. selcnops and O. no-
comis. however, differ markedly from that of other ophidiine
larvae. The possession of an elongate larva is a derived character
uniting these two species.
Robins and Bohike (1959) recognized the close relationship
between O. selenops and O. nocomis based upon the shared
possession of a well-developed rostral spine, similar to that
found in Lepophidium, and the tendency for the dorsal fin to
originate relatively far back on the body. The larvae of Chilara
taylori are slightly more elongate than typical ophidiine larvae,
but bear no close similarity to the larvae of O. selenops.
A character shared by all Lepophidium larvae examined by
Gordon ( 1 982) is the presence of an elongate cartilaginous epural
which ossifies by 15 mm SL. All larvae of the tribe Ophidiini
develop a short cartilaginous epural by 10 mm, but the epural
never ossifies and is not visible by 15 mm SL. The presence of
an epural in the caudal skeleton of the adults is presumably the
primitive character state for the subfamily.
The shared pigmentation pattern oi Lepophidium larvae unites
these species. This character may not extend throughout the
tribe, however. Brownell (1979) illustrates a Genypterus larva
(day eight) that has a pigmentation pattern similar to that of
Lepophidium. Other stages resembled Ophidion, however, with
stellate melanophores scattered laterally on the body. The pos-
session of similarly pigmented larvae by closely related species
in Ophidion argues for the validity of pigmentation as a character
to show phyletic relationships. The larval pigmentation of O.
selenops and O. nocomis shows only slight differences as does
larval pigmentation of O. welshi and O. marginatum. If the
proposed identities of Ophidion Type I and Type 2 and Le-
pophidiumType 1 are correct (Gordon, 1982), then species which
these types represent are presumably closely related.
Adult carapids are morphologically conservative and present
some difficulty in identification and elucidation of phylogenetic
relationships. Larvae, on the other hand, are reasonably well-
known for all genera, fall into fairly distinct morphological groups
and provide morphological diversity which is potentially useful
in understanding intra-familial relationships (Olney and Mar-
kle, 1979; Markle and Olney, 1 980). Robins and Nielsen ( 1 970)
and Cohen and Nielsen (1978) recognized a single family, Carap-
idae, consisting of two subfamilies: the Pyramodontinae with
two genera, Pyramodon and Snyderidia: and the Carapinae with
four genera, Carapus, Echiodon, Encheliophis and Onu.xodon.
However, Gosline ( 1 960b) and Trott (1981) considered the Pyr-
amodontidae a separate family while Arnold (1956) ignored this
GORDON ET AL.: OPHIDIIFORMES
319
group in his revision of carapids. Williams { 1 984) in his synopsis
considered it as a subfamily. The common possession of a vex-
illifer larva is the most convincing evidence that the genera of
Carapidae are monophyletic, thus we recognize one family.
The genera Pyramodon and Snyderidia were considered closely
related by Robins and Nielsen (1970), and Markle and Olney
(1981), on the basis of osteological and larval characters, added
further support to this presumed relationship. It now appears
that many of the character states of these genera are primitive.
The pelvic fins, lost in all other carapids, are obviously a prim-
itive state since they are widely present in all other ophidiiforms.
Similarly, the dorsal origin is over or in advance of the anal fin
in all non-carapid ophidiiforms as well as in Pyramodon and
Snyderidia. The posterior placements of the first dorsal fin ray
or vexillum can therefore be considered advanced states. Thus,
the anterior placement of the vexillum relative to first anal ray
(a primitive state) in combination with a posteriorly placed first
dorsal fin ray (advanced state) appears to define larvae of Car-
apiis (Fig. 162) and presumably Encheliophis. The genera pos-
sess further derived states such as adult inquiline behavior and
parasitism (Trott, 1970). In addition, the tenuis stage, unknown
in Pyramodon and Snyderidia. may represent an advanced state,
namely retaining larval characters in the early benthic stage.
Larvae of the genus Echiodon display a wide variety of mor-
phology especially in gut configuration, vexillum and first dorsal
fin ray position and dorsal pterygiophores (Fig. 161; Maul, 1976;
Olney and Markle, 1979; Markle et al., 1983). Williams and
Shipp (1982) consider Echiodon to be composed of two species
complexes and the gross morphology of larvae seems to support
this contention. In addition, the peculiar gut configuration of
E. rendahli (Fig. 161B, Robertson 1975b) represents another
extreme in morphological variability which suggests the genus
(as presently known) is polyphyletic.
Inter-ordinal relationships. — Based upon anatomical similari-
ties shared with the cods, the ophidiiform fishes have been
treated as a suborder within Gadiformes (Greenwood et al.,
1966; Rosen and Patterson, 1969). These similarities include
the development of the levator maxillae superioris and the struc-
ture of the caudal skeleton. Freihofer (1963, 1970) presented
further evidence for this relationship based upon the pattern of
the ramus lateralis accessorius nerve. Alternatively, these sim-
ilarities may be the result of convergence due to the require-
ments of bottom feeding behavior (Gosline, 1968; Fraser 1972b;
Marshall and Cohen, 1973). Similarities to the perciform fishes
in osteology (Gosline, 1968; Regan, 1912b) and biochemistry
(Shaklee and Whitt, 1981) point to a perciform origin of the
group.
The structure and the development of the ophidiiform caudal
skeleton support the hypothesis of a closer relationship to the
gadiform fishes than to the perciform fishes. In Brotula. as in
gadiforms, two separate ural centra support hypurals. In the
Ophidiinae, a single urostyle, which develops from a single car-
tilaginous structure in the larvae, supports two hypurals. This
urostyle is apparently homologous to the two ural centra of
Brotula. A vestigial neural arch develops on the urostyle, as on
the first ural centrum oi Brotula. Also, the last neural and haemal
spines in both Brotula and the Ophidiinae are modified. These
spines support caudal rays in Brotula and share in the support
of the last dorsal and anal rays in the Ophidiinae. In the gadiform
caudal skeleton, similarly modified spines on the first preural
centrum support caudal rays. In both gadiform fishes (Markle,
1982) and ophidiine fishes these spines remain cartilaginous on
the distal articular surface.
The ophidiine caudal skeleton differs from perciform skele-
tons in the development of the hypural elements and last two
haemal arches. In ophidiine larvae, only two cartilaginous hy-
pural elements form, whereas five or more hypural elements are
typically present in the skeleton of larval perciforms. The last
two haemal arches in perciform fishes remain autogenous; these
arches fuse to the corresponding centra in the Ophidiinae.
Ophidiiform larvae share other developmental features with
gadiform larvae. Larvae of both orders develop coiled guts (ex-
cept for aphyonid larvae) and larvae of Carapidae and Ma-
crouridae have high vertebral numbers resulting in elongate
larvae with reduced or absent caudal fins. Another similarity
apparent in the orders is the presence, in larvae of some species,
of modified anterior dorsal rays. In Ophidiiformes, this char-
acter IS present in larval Carapidae. In Gadiformes, somewhat
similar structures appear in larvae of Bregmaceros, Enchelyopus
and Muraenolepis although comparative studies of the gross
and micro-structure of these larval specializations are lacking
(Govoni et al., 1984).
Cohen and Nielsen (1978) consider ophidiiform fishes to be
too poorly known to resolve questions of phylogeny. Our as-
sessment based on larval data is similar. Further comparative
studies focusing on the developmental osteology of such struc-
tures as the caudal fin, anterior vertebral column and pectoral
girdle, as well as the development of the gut, will allow mean-
ingful interpretation of the significance of these structures to
phyletic studies.
(D.J.G.) RosENSTiEL School of Marine and Atmospheric
Science, University of Miami, 4600 Rickenbacker
Causeway, Miami, Florida 33149; (D.F.M.) The
Huntsman Marine Laboratory, St. Andrews, New
Brunswick EOG 2X0, Canada; (J.E.O.) Virginia Institute
OF Marine Science, College of William and Mary,
Gloucester Point, Virginia 23062.
Lophiiformes: Development and Relationships
T. W. PlETSCH
THE order Lophiiformes is an assemblage of 18 families, 63
genera, and approximately 282 living species of marine
teleosts, the monophyletic origin of which seems certain based
on the following synapomorphic features: (1) Spinous dorsal fin
primitively of six spines, the anteriormost three of which are
cephalic in position and modified to serve as a luring apparatus
[involving numerous associated specializations, e.g., a medial
depression of the anterior portion of the cranium, loss of the
nasal bones (nasal of Rosen and Patterson, 1969 = lateral eth-
moid) and supraoccipital lateral-line commissure, and modifi-
cations of associated musculature and innervation]; (2) Epiotics
separated from parietals and meeting on the midline posterior
to the supraoccipital; (3) Gill opening restricted to a small, elon-
gate tubelike opening situated immediately dorsal to, posterior
to, or ventral to (rarely partly anterior to) pectoral-fin base; (4)
Second ural centrum fused with the first ural and first preural
centra to form a single hypural plate (sometimes deeply notched
posteriorly) that emanates from a single, complex half-centrum
(Rosen and Patterson, 1969:441, text figs. 4E, 60); (5) Pectoral
radials narrow and elongate, the ventral-most radial consider-
ably expanded distally; and (6) Eggs spawned in a double, scroll-
shaped mucous sheath (Rasquin, 1958).
Within the order there are currently recognized three subor-
ders: the Lophioidei, containing a single family and 25 species
of relatively shallow- water, dorso-ventrally flattened forms (Ca-
ruso, 1981, 1983; Caruso and Bullis, 1976); the Antennarioidei,
with six families and approximately 121 species, nearly all lat-
erally-compressed, shallow-water, benthic forms (Bradbury,
1967; Pietsch, 1981, 1984; Pietsch and Grobecker, in press);
and the Ceratioidei, containing about 136, typically globose,
meso- and bathypelagic species (Bertelsen, 1951; see also Ber-
telsen, this volume).
Development
Little is known about the early life stages of lophiiform fishes,
unequal information being available for only the Lophiidae,
Antennariidae and most ceratioid families. Eggs are well-known
in lophiids (Fulton, 1898; Bowman, 1920; Bigelow and Welsh,
1925) and antennariids(Mosher, 1954; Rasquin, 1958) but un-
known in all other lophiiforms. Larvae are adequately described
in lophiids (Bowman, 1920; Martin and Drewry, 1978), anten-
nariids(Mosher, 1954; Rasquin, 1958) and most ceratioids (Ber-
telsen, 1951), but remain undescribed in chaunacids and ogco-
cephalids despite some available material.
Probably the most striking characteristic of early ontogeny in
lophiiforms is the fact that eggs are spawned embedded in a
continuous, ribbon-like sheath of gelatinous mucous, often re-
ferred to as an "egg-raft" or "veil" (with one known exception,
see Pietsch and Grobecker, 1980). Within this mucous veil are
found thousands of roughly-hexagonal, liquid-filled chambers
arranged in one to several irregular layers, each chamber con-
taining from one to three eggs (see Rasquin, 1958 for further
details and figures of the structure of egg rafts). Development
is fairly direct, with the larvae in all known groups gradually
acquiring adult characters over a size range of approximately 5
mm total length (TL) in antennariids to 65 or 70 mm TL in
lophiids. Specialized ontogenetic stages are absent except for the
peculiar "scutatus" prejuvenile present in the ontogeny o{ An-
lennarius radiosus (see below).
Lophiidae
Of the 25 species and four genera of the Lophiidae (Caruso,
198 1), early life stages have been described for only three species,
all of the genus Lophius: L. americamts (Martin and Drewry,
1978, and numerous references cited therein), L. piscatohus
(Tuning, 1923) and L. /'(/a'e^a5M(Padoa, 1 956e). Of these, early
ontogeny is best documented in L. americanus. a spring or
summer spawner, whose egg rafts measure 0. 1 5-1 .5 m wide and
6-12 m long. Living eggs are slightly oval, their major axis
measuring 1.61-1.94 mm. The outer shell appears smooth and
transparent, the yolk homogeneous and amber in coloration.
The perivitelline space is narrow. A single, copper, orange or
pinkish-colored oil globule is present, having a diameter of ap-
proximately 0.40-0.45 mm. Yolk-sac larvae measure 2.5-4.9
mm TL. The larvae, ranging in size from 6.5 to approximately
10.5 mm TL, are prominently pigmented, with early-forming
dorsal rays and pectoral and pelvic fins (Fig. 164A). Relative to
antennariid larvae, the head is small, somewhat less than 30%
of standard length. The gut is unusually short. The dorsal and
pelvic rays are unusually elongate. The soft-dorsal and anal fins
are last to form. The pectoral fin is typically large and fan-
shaped. Fin ray counts are complete by approximately 12 mm
TL. Transformation to the prejuvenile stage takes place at a size
somewhat greater than approximately 12 mm TL; the juvenile
stage is not reached until at least 65 mm TL. In well preserved
specimens of some species (i.e., Lophiodes spilurus; SIO 59-324,
65.5 mm TL) the epidermal layer of the head and body is greatly
distended by transparent, gelatinous connective tissue, giving
the larvae an inflated or balloon-like appearance (as described
for ceratioid larvae by Bertelsen, 1951:12; see also Bertelsen,
this volume). (Largely taken from Martin and Drewry, 1978:
359-366, where the reader will find a full series of figures and
more detailed description of early ontogeny.)
Although the significance of variation in larval pigmentation
in lophiids is largely unknown, larvae of the American species,
Lophius americanus Valenciennes, are more easily distin-
guished from those of the European L. piscatorius Linnaeus than
are the adults, using characteristic differences in pigmentation
(Taning, 1923; Martin and Drewry, 1978). Tuning (1923), after
studying early developmental stages, considered the two species
to be distinct at a time when many authors regarded them as
synonyms (Martin and Drewry, 1978).
Meristic characters that typify early life stages of lophiids are
compared with those of other lophiiforms in Table 88.
Antennariidae
The family Antennariidae consists of 41 species distributed
among 1 2 genera (a modification of Schultz, 1957; Pietsch, 1981,
1 984; Pietsch and Grobecker, in press). Of these, early life stages
320
PIETSCH: LOPHIIFORMES
321
»i>rVf-^--^'
Fig. 164. Representative larvae of lophiiform fishes: (A) Lophius americanus. 12 mm TL, taken from Martin and Drewry, 1978:364, fig.
191B; (B) Hislrio hislrio. 5.3 mm TL, taken from Adams, 1960:64, fig. IB; (C) Chaunax sp., 9.8 mm TL, ZMUC P922155. Gulf of Mexico,
22°06'N, 84°58'W; (D) Ogcocephalus sp.. 10.4 mm TL, SHL D-66-12, P-4, Western North Atlantic. 34<'17'N, 76°23.5'W; (E) Caulophrvnejordani.
9.5 mm TL, ZMUC P92198. taken from Bertelsen, 1951:35, fig. UB.
have been described in Antennarius stria! us (Phrynelox scaber
and P. nuttingi oi Kas(\\i\n, 1958, and P. sca/)er of Martin and
Drewry. 1978) and Histrio histrio (Martin and Drewry, 1978
and numerous references cited therein), with only brief descrip-
tions of the "scutatus" prejuvenile stage of A. radiosus (see
below). Of these, H. histrio is by far the best known. Spawning
occurs year-round except in February and March. Freshly
spawned egg rafts measure approximately 25-50 mm wide and
90 mm long. Eggs are initially oval in shape (their major axis
measuring 0.62-0.65 mm), but become spherical at the time of
the second cleavage. Ova are extremely transparent (Mosher,
1954) and colorless, without oil globules. As development pro-
ceeds, the raft unrolls, expanding to a length of 90 cm (Smith,
1907). The membranes remain firm until about the 6-1 1 myo-
mere stage, but then begin to deteriorate, the raft softening and
expanding to about three times its original dimensions, and
finally beginning to sink. Yolk-sac larvae measure 0.88-1 .7 mm
TL. The larvae, most strikingly characterized by their large head
(greater than 45% standard length), range in size from approx-
imately 1.6-7.2 mm TL (Fig. 164B). Pigmentation is conspic-
uous about the head and midgut. The base of the pelvic fin
elongates at about 12 mm TL, at which time the pigmentation
of the midgut begins to fade. The sequence of fin formation is
as follows: caudal, anal, soft-dorsal, pelvic, pectoral, the dorsal
spines being the last to form at approximately 13 mm TL.
Prejuveniles range in size from approximately 7,3-20 mm TL.
(Taken from Martin and Drewry, 1978:372-384, where the
reader will find a full series of figures, and a more detailed
description of the early development of H. histrio, as well as
that o{ A. striatus.)
322
ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM
Table 88. Meristic Values for Major Taxa of the Lophiiformes.
Character
Lophiidae
Antenna riidae
Chaunacidae
Ogcocephalidae
Ceralioidei
Dorsal fin
II-III + O-III + 8-
-12
III + 10-16
III + 10-12'
II + 1-6=
II + 3-22=
Anal fin
6-10
6-10
6-7
3-4
3-19
Caudal fin
8
9
8
9
8-10
Pectoral fin
13-28
6-14
11-14
10-15
12-30
Pelvic fin
I + 5
I + 5
I + 4
1 + 5
3-4'
Branchiostegal
rays
2 + 4
2 + 4
2 + 4
2 + 4
1 + 4, 2 + 4
Vertebrae
19-31
18-23
19
18-21
18-24
' Second and third spine embedded beneath skin of head-
' Second spine reduced to a tiny remnant and embedded beneath skin of head.
* Present only in larvae of the ceratioid family Cauiophrynidae.
Meristic characters that typify early life stages of antennariids
are compared with those of other lophiiforms in Table 88.
The so-called "scutatus" prejuvenile form, originally de-
scribed as a new genus and species, Kanazawaichthys scutatus,
by Schultz (1957), but later found by Hubbs (1958) to be the
prejuvenile of Antennarius radiosus, remains unique (Fig. 165).
The primary morphological features that characterize these early
life stages are so drastic (a pair of shield-like, bony extensions
of the cranium that reach posteriorly beyond the level of the
opercular bones, and an expansion of the anterior margin of the
bones of the suspensorium; see Schultz 1957:63, plate 14, fig.
A, and Hubbs, 1958) that their appearance in other antennariids
of similar sizes, particularly among closely related species (such
as A. ocellatus and A. avalonis), is to be expected. Yet, despite
the fact that numerous other species are represented by small
specimens, no comparable morphological adaptations have been
discovered.
Chaunacidae
The family Chaunacidae contains a single genus and as many
as 12 species (J. H. Caruso, pers. comm., 8 June 1983). Aside
from the fact that larvae and "young specimens" are often caught
bathypelagically (Mead et al., 1 964), nothing has been published
on their early life stages, despite some available material. The
ovaries of members of this family are scrolled like those of other
lophiiforms, suggesting the production of epipelagic egg rafts,
although neither eggs nor rafts have been reported.
The material of Chaunax available to me, 32 specimens in
25 lots (all part of the DANA collections of the Zoological
Museum, University of Copenhagen), measured 4.3-10.6 mm
TL. Even the smallest of these appear to have reached a pre-
juvenile stage, with all fin rays formed (including the illicium),
and the skin everywhere covered with close-set dermal spinules
(Fig. I64C). Pigmentation appears to be absent. In well pre-
served specimens, the epidermal layer of the head and body is
greatly distended by transparent, gelatinous connective tissue.
As in antennariids, the head is large, considerably greater than
50% of SL in all specimens examined.
Meristic characters that typify the early life stages of chaun-
acids are summarized and compared to those of other lophii-
forms in Table 88.
Ogcocephalidae
The Ogcocephalidae contains nine genera and approximately
60 species (Bradbury, 1967). Like chaunacids, practically noth-
ing is known about their early life stages. The only published
information, aside from a report of the capture of a single, 17.5
mm TL specimen of Ogcocephalus sp. by Clark et al. (1969), is
a comment by Mead et al. (1964) that larval and postlarval
specimens have been caught epipelagically. The scrolled ovaries
of members of this family may indicate that egg rafts are pro-
duced, but neither eggs nor rafts have been reported.
The larval ogcocephalid material available to me (29 speci-
mens in nine lots, kindly provided by Michael P. Fahay of the
Northeast Fisheries Center, Sandy Hook Laboratory, and ten-
tatively identified as Ogcocephalus sp.) measured 3.1-18.4 mm
TL. In the smallest of these, all fins are fully developed, except
for the illicium; a tiny rudiment of this structure is just barely
visible in a 4.9 mm TL specimen, but relatively conspicuous in
a 5.1 mm TL specimen (the transition to a prejuvenile stage is
thus taken to occur at approximately 5.0 mm TL). By a length
of 8.6 mm TL the pectoral fins are large and fan-like, the base
of the pelvic fin has become elongate, and small, scattered me-
lanophores are present on top of the head, nape of the neck, on
the cheek just behind the eye, the pectoral fin base and blade,
and on the caudal peduncle (Fig. 164D). At 8.9 mm TL the
pigmentation is well developed, and the paired fins are dispro-
portionately large. At 12.4 mm TL dermal spinules are begin-
ning to form in the skin; a lateral, longitudinal cluster of dermal
spinules, which will later form the ridge-like, outermost margin
of the adult, is just beginning to develop. By 18.4 mm TL the
skin is everywhere covered with broad-based dermal spinules,
and the lateral ridge is well-developed. At all stages of devel-
opment, but particularly the prejuvenile stage, the skin is highly
inflated, giving the larvae an almost spherical shape. At all stages
the head is disproportionately large, considerably greater than
50% of standard length.
Meristic characters that typify the early life stages of ogco-
cephalids are summarized and compared to those of other lo-
phiiforms in Table 88.
Ceratioid Families
The Ceratioidei contains 1 1 families, 34 genera and approx-
imately 1 36 species. Isolated eggs of ceratioids are unknown;
ovarian eggs, described in only a few species, are slightly oval,
the major axis of the largest of these measuring 0.50-0.75 mm.
The larval stages of all of the families and most of the genera
have been described (Bertelsen, 1951). Generally speaking, cer-
atioid larvae are typically small. According to Bertelsen (1951),
the smallest known larvae measure 2.0-3.0 mm TL, whereas
the largest larvae and smallest metamorphosis stage range from
12-25 mm TL for females, and 10-22 mm TL for males. As in
antennariid, chaunacid and ogcocephalid larvae, the head is
disproportionately large; as in some lophiids, chaunacids and
ogcocephalids, the head and body are enveloped by transparent.
PIETSCH: LOPHIIFORMES
323
Fig. 165. "Scutatus" prejuvenile oi Amennarius radiosus, 21.2 mm XL, USNM 251937-F21. North Atlantic, 36°30'N, 74°30'W; drawn by
B. Washington.
highly-inflated skin (Fig. 164E; for details see Bertelsen, 1951;
Bertelsen, this volume).
Meristic characters that typify the early life stages of ceratioids
are summarized for all eleven families and compared to those
of other lophiiforms in Table 88.
Relationships
Since Regan (1912a), three major lophiiform taxa of equal
rank have been recognized by nearly all authors. These taxa,
together with their currently recognized families (the eleven
families of the bathypelagic Ceratioidei excluded; see Bertelsen,
this volume), are: Suborder Lophioidei — Family Lophiidae;
Suborder Antennarioidei — Families Antennariidae, Tetrabra-
chiidae, Lophichthyidae, Brachionichthyidae, Chaunacidae, and
Ogcocephalidae; Suborder Ceratioidei. ,
On the basis of adult characters alone, Pietsch (1981:416, fig.
41) attempted to test the validity of Regan's (1912a) three
major lophiiform taxa using cladistic methodology, but ran into
serious difficulty in attempting to establish monophyly for the
Antennarioidei; while a number of synapomorphic features were
found to support a sister-group relationship between the four
families Antennariidae through Brachionichthyidae, and be-
tween the families Chaunacidae and Ogcocephalidae, no con-
vincing synapomorphy was found to link these two larger
subgroups.
To date, early life history stages have not been used in for-
mulating hypotheses of relationship among lophiiform fishes;
but, several egg and larval characters are shown here to be
significant in resolving a number of problems with this group.
These characters, along with several previously unidentified adult
characters, have been used here to construct a revised cladogram
of lophiiform relationships (Fig. 166).
This new cladogram differs significantly from that previously
published (Pietsch, 1981: fig. 4 1 ). The suborder Antennarioidei
is now restricted to only four families: The Antennariidae, rec-
ognized as the primitive sister-group of the Tetrabrachiidae,
these two families together forming the primitive sister-group
of the Lophichthyidae, and this assemblage of three families
forming the primitive sister-group of the Brachionichthyidae.
These relationships are supported by a total of seven synapo-
morphies (most of which were previously described by Pietsch,
1981:413-414) numbered 7 through 13 in Fig. 166: (7) Pos-
teromedial process of vomer emerging from ventral surface as
a laterally-compressed, keel-like structure, its ventral margin (as
seen in lateral view) strongly convex (Pietsch, 1981:397, figs.
4-6); (8) Postmaxillary process of premaxilla spatulate (Pietsch,
1981:398, figs. 8, 20); (9) Opercle similarly reduced in size
(Pietsch, 1981:401, figs. 9, 21); (10) Ectoplerygoid triradiate, a
dorsal process overlapping the medial surface of the metapter-
ygoid (Pietsch. 1981:400, figs. 9, 21, 22); (11) Proximal end of
hypobranchials II and III bifurcated (Pietsch, 1981:407, figs. 11,
28. 29); (12) Interhyal with a medial, posterolaterally directed
process that comes into contact with the respective preopercle
(Pietsch, 1981:400, fig. 26); and (13) Illicial pterygiophore and
pterygiophore of the third dorsal spine with highly compressed,
blade-like dorsal expansions (Pietsch, 1981:410, figs. 36, 37).
The present interpretation of lophiiform relationships differs
further from any previously proposed hypothesis in considering
the Antennarioidei (sensu stricto) to form the primitive sister-
group of a much larger group that includes the Chaunacioidei
(new suborder), the Ogcocephalioidei (new suborder) and the
Ceratioidei. The Ogcocephalioidei is, in turn, recognized as the
primitive sister-group of the Ceratioidei (Fig. 166).
Monophyly for a group containing the suborders Antenna-
rioidei, Chaunacioidei. Ogcocephalioidei and Ceratioidei is sup-
ported by four, previously unidentified synapomorphies (num-
bered as they appear in Fig. 1 66): ( 1 4) Eggs and larvae small (at
all stages eggs are considerably less than 50% the diameter of
324
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 166. Cladogram showing proposed phylogenetic relationships of the major subgroups of the Lophiiformes. Black bars and numbers refer
to synapomorphic features discussed in the text.
those of lophioids; smallest larvae are certainly less than 50%,
and probably less than 30% the size of those of lophioids; size
at transformation to the prejuvenile stage is less than 60% that
of lophioids); (15) Head of larvae proportionately large rel-
ative to body (always greater than 45% of standard length, com-
pared to less than 30% in lophioids); (16) Reduction in the
number of dorsal fin spines from a primitive number of six in
lophioids to three or less (Pietsch, 1981:409, figs. 36-38); and
(17) Loss of pharyngobranchial IV (present and well-toothed in
lophioids; Pietsch, 1981:401, figs. 1 1, 28-32).
Monophyly for a group containing the suborders Chauna-
cioidei, Ogcocephalioidei and Ceratioidei is supported by two
synapomorphies: (18) Second dorsal spine reduced and embed-
ded beneath skin of head (Pietsch, 1981:410, figs. 36-38); and
(19) Gill filaments of gill arch I absent (but present on proximal
end of ceratobranchial I of some ceratioids; Bradbury, 1967:
408; Pietsch, 1981:415).
That the Ogcocephalioidei is the primitive sister-group of all
ceratioid families is supported by three synapomorphies: (20)
Second dorsal spine reduced to a small remnant (well developed
in the ceratioid family Diceratiidae. and in all other lophiiforms;
Bertelsen, 1951:17; Pietsch, 1 98 1 :4 1 0, fig. 38); (2 1 ) Third dorsal
spine and pterygiophore absent (present in all other lophiiforms;
Bertelsen, 1951:17; Bradbury, 1967:401; Pietsch, 1981:410, fig.
38); and (22) Epibranchial I simple, without ligamentous con-
nection to epibranchial II (in batrachoidiforms and all other
lophiiforms epibranchial I bears a medial process that is liga-
mentously attached to the proximal tip of epibranchial II; Pietsch,
1981:401, figs. 28-32).
Of the possible cladograms that could be constructed on the
basis of the data provided in this study, the one shown in Fig.
166 is by far the most parsimonious. But at the same time,
acceptance of this revised hypothesis of relationships of lo-
phiiform fishes requires evolutionary convergence or reversal
in three derived character states previously used by me (Pietsch,
1981:415, fig. 41) to support a hypothesis of sister-group rela-
tionship between the Chaunacidae and Ogcocephalidae: ( 1 ) Pos-
teriormost branchiostegal ray exceptionally large (all four pos-
teriormost branchiostegal rays approximately equal in size in
batrachoidiforms and all other lophiiforms; Pietsch, 1981, fig.
27); (2) Gill teeth tiny, arranged in a tight cluster at apex of
pedicel-like tooth plates (in all other lophiiforms gill teeth, if
present, are relatively large, and either single, or associated with
a flat, rounded tooth plate; but tiny, and at apex of elongate
pedicel-like tooth plates in at least some batrachoidiforms, e.g.,
Poriclithys; Pietsch, 1981, figs. 31,32) and (3) Illicial bone, when
retracted, lying within an illicial cavity (an illicial cavity is absent
in all other lophiiforms; however, the illicium and esca lie within
a shallow groove on the dorsal midline, sometimes enveloped
by folds of skin, in the antennariid genus Histiophryne, Pietsch,
1981, fig. 39; Pietsch, 1984:40).
The cladistic relationships of the Lophiiformes are summa-
rized in the following revised classification. While the ranking
of taxa is not dichotomous (see methods in Pietsch, 1981:388),
PIETSCH: LOPHIIFORMES
325
inter-nested sets of vertical lines are used to indicate mono-
phyletic units.
Order Lophiiformes
Suborder Lophioidei
Family Lophiidae
Suborder Antennarioidei
Family Antennariidae
Family Tetrabrachiidae
. Family Lophichthyidae
Family Brachionichthyidae
Suborder Chaunacioidei New
Family Chaunacidae
Suborder Ogcocephalioidei New
Family Ogcocephalidae
Suborder Ceratioidei
As a final note, the Lophiiformes has traditionally been allied
with the Batrachoidiformes, based primarily on osteological
characters ofthe cranium (Regan, 1 9 1 2a; Gregory, 1933; Rosen
and Patterson, 1969). However, this sister-group relationship
has yet to be shown conclusively, and 1 have not been able to
assess the significance of early life stages in supporting or refuting
this hypothesis.
School of Fisheries WH- 1 0, College of Ocean and Fisheries
Sciences, University of Washington, Seattle,
Washington 98195.
Ceratioidei: Development and Relationships
E. Bertelsen
THE Ceratioidei differ most distinctly from all other mem-
bers ofthe order Lophiiformes in being meso- and bathy-
pelagic, lacking pelvic fins (except in larval Caulophrynidae)
and in having extreme sexual dimorphism. Males are dwarfed
and differ from females in lacking an external illicium and hav-
ing denticular teeth on the tips of the jaws and well-developed
eyes and/or olfactory organs. Furthermore, Ceratioidei differ
from other Lophiiformes, except the family Ogcocephalidae, in
lacking a third cephalic ray and its pterygiophore, and except
in the families Caulophrynidae, Neoceratiidae and the gigan-
tactinid genus Rhynchactis. females ofthe suborder differ from
all other Lophiiformes in having a bulbous swelling of the tip
of the illicium (escal bulb) containing a large globular photo-
phore.
The suborder contains approximately 1 34 species placed in
34 genera and 1 1 families (Table 89). The taxonomy is based
mainly on studies of the females. Except for the larval stages
and the basic meristic and osteological characters shared by the
two sexes, descriptions require separate treatment of females
and males.
The separation into families is based mainly on osteological
characters, of which some ofthe more important are compared
in Table 89. Most ofthe families form well-defined and mutually
very distinct taxa in which the females (especially) possess unique
morphological features which separate them from members of
all other families.
The separation into genera is based mainly on characters
present only in females. Somewhat varying between families,
some ofthe most important of these characters are differences
in: (1) shape of skull and other bones of the head including
development of its spines; (2) jaw mechanisms, including den-
tition; (3) illicial apparatus, including basic patterns of escal
appendages; and (4) pigmentation of skin and development of
dermal spines. Some of the distinguishing osteological charac-
ters, especially in shape of opercular bones, are shared with the
males, like the fin ray numbers which in some families show
distinct intergeneric differences. The special male structures,
such as denticular teeth, show distinct intrageneric differences
in full agreement with the separations based on characters of
the females.
The species of Linophryne have been grouped into subgenera
and those of Himantolophus (in ms.), Oneirodes. and Gigan-
tactis into "species-groups'" based on shared minor differences
in one or more of the characters mentioned above. All intra-
generic separations of the females into species are based on
differences in pattern and shape of escal appendages, often com-
bined with differences in illicial lengths. In a majority of the
recognized species, no other separating characters have been
shown. In others, differences in meristic characters (numbers of
fin rays and teeth) and minor osteological characters (shape of
opercular bones, development and dentition of branchial arches,
etc.) have been observed supporting the separation into species.
A special opportunity to check the validity of the separations
based on escal characters is found in the genus Linophryne. in
which females have hyoid barbels which in pattern of branching
show very distinct differences between species and subgenera
(Bertelsen, 1982).
In most cases it has not been possible to separate males into
taxa below the level of genera and subgenera. A few males
differing from their supposed congeners in special male char-
acters (especially denticular teeth) have been tentatively de-
scribed as representatives of separate species. Studies of males
attached to identified females have not revealed characters of
specific order.
Development
No spawned, fertilized eggs of Ceratioidei have been described
(re-examination by Bertelsen, 1980:66, of an egg referred to
Linophryne arborifera by Beebe, 1932:93, indicated that it rep-
resents a diodontid).
326
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 89. Characteristics of Ceratioidei.
X: presumed denved characters
O: presumed pnmilivc characters
Caul-
Nco-
Thauma-
Centro-
Lin-
ophryni-
cerati-
Melano-
Himantol-
Dicerali-
Oneir-
hchthyi-
phryni-
Cerali-
Gigantac-
ophryn-
Families (11):
dae
idae
cetidae
ophidae
idae
odidae
dae
dae
idae
tinidae
idae
No. of
genera (34):
2
1
I
1
2
15
2
1
2
2
5
No. of species (ca. 134):
4
1
4
ca. 15
4
52
6
1
3
19
25
No.
Characters shared by sexes:
1
Dorsal finrays
6-22
11-13
12-17
5-6
5-7
4-8
5-7
6-7
4
3-9
3
2
Anal finrays
5-19
10-13
4
4
4
4-7
4-5
5-6
4
3-7
3
3
Caudal finrays
8
9-10
9
9
9
9
9
9
8-8 Vj
8'/2
8 1/2
4
Branchiostegal rays
5-6
5-6
6
6
6
6
6
6
6
6
5
5
Pectoral radials
2
3
4
3
3
3
3
4
4
5
3
6
Shape of pelvic bones
XI
XI
O
O
O
O-Xl
O
XI
XI
Xl-2
X2
7
Head of hyomandibular
o
X
O
O
O
O-X
O
O
O
X
X
8
Parietals
o
O
o
X
o
O
O
O
o
o-x
O
9
Pterosphenoid
X
O
o
o
o
O-X
o
O
X
X
X
10
Epural
Female characters:
o
X
X
X
X
X
X
X
X
X
O-X
11
Escal photophore
o
O
X
X
X
X
X
X
X
O-X
X
12
Photophore on 2'ccphalic ray
o
O
o
o
X
O
o
O
X
o
O
13
Separation of frontals
o
O
X
X
X
X
X
X
o
X
O-X
14
Shape of frontals
o
O
o
o
o
X
X
X
X
XI
X
15
Maxillaries
XI
XI
o
o
o
o
o
O
o
X2
Xl-2
16
Maxillomandibular ligament
X
X
o
o
o
o
X
o
o
X
X
17
Branchial teeth
X
X
X
o
X
o-x
X
o
X
X
O-X
18
Quadrate and Articular spine
o
O
o
o
o
O-X
X
o
o
o
o
19
Dermal spines
X
X
o
o
o
O-X
O-X
o
o
o
X
20
Shape of body
Male characters:
o
X
o
o
o
o
X
X
X
X
o
21
Eyes
o
XI
o
o
o
o
o
XI
X3
XI
X2
22
Olfactory organs
X
o
X
X
X
X
X
X
o
X
X
23
Anterior nostrils
X
o
o
o
o
X
X
X
o
X
X
24
Upper denticular teeth
o
o
X
X
X
X
X
X
X
X
X
25
Dermal spines
X
X
o
o
o
X
o
X
o
O-X
X
26
Parasitic males observed
Larval characters:
X
X
o
o
o
X
o
o
X
o
X
27
lllicial rudiment
o
o
X
X
X
X
X
X
X
X
X
28
Pelvic fins
o
X
X
X
X
X
X
X
X
X
X
29
Pectoral fins
o
X
X
X
X
X
X
X
X
o
X
30
Shape of body
o
X2
o
o
o
O-Xl
o
o
X3
o
XI
(1-2) High numbers of dorsal and anal fin rays possibly a pnmitive character state; (3) Nine rays presumed pnmitive; 8'/:: 9th ray reduced to less than half length of the 8th. (4) Six rays
presumed pnmitive; (5) Three radials. shared with antennanoids, here presumed to be pnmitive; however, the trend to reduce the number from 4 to 3 in older specimens of centrophrymds and
melanocetids might indicate that within ceratioids four radials are pnmitive and three a result of secondary reduction; (6) Distally expanded pelvic bones (triradiate in Himanfolophus and some
specimens of the oneirodid genus Chaenophryne). presumed pnmitive; XI: a simple rod; X2: absent; (7) Double head of hyomandibular presumed pnmitive; X: single head (in oneirodids in
only one of the 15 genera); (8) Panetals absent in all himantolophids. lost in metamorphosed females of the gigantactinid genus Rhynchactis. (9) Presence of pterosphenoid presumed pnmitive
(absent in one of the oneirodid genera); (10) A single epural observed in Caulophn'iie and in the linophrynid genus Phnlocon'nus. (11) The absence of escal photophore presumed pnmitive in
Caulophrynidae. and possibly in Neoceraltas while the absence in the gigantactinid genus Rhynchaclis is presumed to be due to a secondary specialization (cf text); (12) While the presence of a
photophore on 2nd cephalic ray is a denved character stale, the presence of this ray may be regarded as primitive; (13) Frontals meeting in the midline presumed pnmitive (in linophrynids
present in Pholocorynus); (26) Parasitic males observed in Neoceraltas, both genera of Ceratiidae, one of the two genera of Caulophrynidae. four of the five genera of Linophrynidae. and one of
the 15 genera of Oneirodidae; (27) Presence of an external illicial rudiment in larval males presumed primitive; (28) Presence of pelvic fins presumed pnmitive, (29) Enlarged pectoral fins here
presumed pnmitive; and (30) Short, more or less sphencal body presumed pnmitive; XI: moderately elongate; X2: slender; X3: hump-backed.
A cluster of eggs embedded in a mucoid substance hanging
out of the greatly expanded genital opening of a sexually par-
asitized female of Ltnophryne arborifera was observed by Ber-
telsen ( 1 980:66). This indicates that Ceratioidei expel their eggs
in free-floating mucoid egg "rafts" or "veils," as described in
species of the other suborders of Lophiiformes (Rasquin, 1 958).
It is possible that the release of the egg veil of the specimen was
caused by the catch, and that the eggs were not completely
mature. They were slightly oval, 0.6-0.8 mm in diameter, with
smooth, very soft outer membranes that were folded or shrunk-
en in several. The yolk, which contained numerous small oil
globules, was opaque and partially surrounded by an irregular
perivitelline space.
The observed number of ceratioid females with apparently
nearly mature ovaries is relatively small. In these, the largest
eggs have diameters of 0.5-0.75 mm.
Larvae and/or metamorphic stages representing all 1 1 fami-
lies and 26 of the 34 recognized genera have been described,
the majority by Bertelsen (1951). Identification to species is
restricted, however, to those genera, subgenera or "species-
groups" in which only a single species is recognized.
No specialized ontogenetic stage between larvae and juveniles
occurs. In most genera, metamorphosis begins at a size of 8-10
mm SL, while in some (Himantolophidae, Thaumatichthys, Gi-
gantactinidae, and Linophrynidae), the larvae may reach lengths
of 15-25 mm. During metamorphosis, covering a size range
somewhat varying between genera, adult characters are gradu-
ally acquired. In both sexes, the skin is gradually covered with
pigment (except in Haplophryne), and in certain genera, skin
spines are developed. In females the illicial and escal characters
develop, the head and especially the jaws increase in relative
size, the larval teeth are replaced, and the growth of eyes and
BERTELSEN: CERATIOIDEI
327
olfactory organs is retarded. In males the body elongates, larval
teeth are lost, the denticular teeth develop, and eyes and/or
olfactory organs increase in relative size.
The larvae have been referred to genera, subgenera or species-
groups on basis of (1) meristic characters (especially number of
dorsal and anal fin rays); (2) osteological characters (especially
number of branchiostegal rays and pectoral radials and shape
of head of hymandibular, pelvic bones and opercular bones);
and (3) pattern of subdermal pigmentation. The pattern is re-
tained under the pigmented skin of post-metamorphic juveniles
which have acquired adult characters (Bertelsen, 1951). In most
genera the smallest larvae observed are 2.5-3.5 mm. At these
stages, in which no distinguishing characters other than pig-
mentation may be developed, identification is based on com-
parison with developmental series of older larvae.
Meristic characters.— The 2-3 mm smallest known larvae have
an almost straight notochord and almost undifferentiated fins.
The fin rays of the unpaired fins are laid down early and the
full number is usually present in larvae of 3-4 mm SL of the
numerous species where the number of dorsal rays does not
exceed 8. The pectoral fin rays are laid down somewhat later
than those of the unpaired fins, and the lowermost rays are rarely
discernible in specimens of less than about 5-6 mm. Caulo-
phrynidae and the ceratiid genus Cryptopsaras have 8 caudal
fin rays, all others have 9 ( 10 in some specimens of Neoceratias).
The 9th (lowermost) ray is rudimentary or short (less than half
the length of the 8th ray) in Linophrynidae, Gigantactinidae,
and Ceratias.
Except in the three genera in which the number of rays in the
anal and/or dorsal fin exceeds 10 (Caulophryne. Neoceratias.
and Melanocelus, cf Table 89), the intraspecific variation of the
number of fin rays in these fins is small, rarely more than ± 1 .
Significant differences in numbers of dorsal and anal fin rays
have been found between species within the genera Caulo-
phryne. Gigantactis, and Melanocetus and between genera in
the families Caulophrynidae, Gigantactinidae, and Oneirodi-
dae.
Pectoral fin rays number 12-23 in all ceratioids (except Cten-
ochirichthys with 28-30). As an intraspecific range of variation
of 5 to 7 fin rays has been observed, this character may aid
identification only in exceptional cases.
All reported vertebral counts of Ceratioidei fall within the
range of 19 to 24, the highest number in Neoceratiidae, the
lowest in Linophrynidae. The limited number of observations
does not permit an evaluation of the diagnostic value of differ-
ences within this range.
Morphology.— The head and body of larval Ceratioidei are sur-
rounded by inflated transparent skin. Due to this balloon-like
envelope, their shape varies from nearly spherical, with greatest
width and depth of body reaching 80-90% SL, to elongated or
pear-shaped, with body depth of 40-60% SL.
The inflation of the skin varies with preservation, but gen-
erally its greatest development is found in Caulophrynidae (Fig.
167 A), Gigantactinidae (Fig. 168A, B), and Himantolophidae
(Fig. 169 A, B); less pronounced in Neoceratiidae (Fig. 167B),
Ceratiidae (Fig. 168C-E) and Oneirodidae (Fig. 170). No dis-
tinct change has been observed in the relative development of
the inflation during larval life. In larvae of most genera, the
relative length of head, measured to base of pectoral peduncle,
is about 50% SL. In Oneirodidae and Linophrynidae it is some-
what less (generally about 45%) and is shortest in Neoceratias
(35-40%). In the late larval stages of males, the elongation of
the body may start before other metamorphic characters have
appeared.
In larval Ceratiidae, the vertebral column is more strongly S-
shaped than in other families, resulting in a characteristic hump-
backed appearance of these larvae (Fig. 168C-E). The larvae of
Caulophrynidae (Fig. 167 A) and Gigantactinidae (Fig. 168A, B)
differ distinctly from those of other Ceratioidei in the size of
the pectoral fins, which have lengths of 40% to nearly 60% SL,
measured from base of pectoral peduncle. In the other families,
this length is generally 20-25% SL and does not exceed 30%
SL.
Pigment.— The subdermal pigment occurring in larval Cera-
tioidei is usually separated into four more or less well-defined
main groups: (I) peritoneal; (2) opercular; (3) dorsal; and (4)
caudal-peduncular.
In Neoceratias (Fig. 167B) and some linophrynids (Haplo-
phryne. Fig. 167D), the subgenera of Linophryne: Stephano-
phryne {Fig. 167F). and Rhizophryne (Fig. 167C), the subdermal
pigment forms dorso- and ventrolateral bands along the body.
In all species in which one of these main groups occur, they
are generally laid down in the youngest larvae as a few small
and scattered melanophores which during larval development
gradually increase in size, number, and in area covered.
Additional groups of melanophores occur in some taxa (for
instance on base of pectoral peduncle in Rhynchactis, Fig. I68B;
internally in fin rays of Pentherichthys, Fig. 1 70E; on the pos-
terior angle of lower jaw in Stephanophryne, Fig. I67F; and on
a swelling of the outer transparent skin in front of the dorsal fin
in some Himantolophus. Fig. I69B).
Complete lack of pigment is found in Ceratias (Fig. 168E),
some Gigantactis, and the linophrynid genus Borophryne (Fig.
167G). Besides in these, peritoneal pigment is absent only in
Neoceratias. In all others, this group is laid down on the dorsal
side of the peritoneum of the youngest larvae and with growth,
gradually spreads to its lateral and posterior sides.
Pigmentation of the opercular region varies greatly between
taxa. It is absent or weakly developed in most genera, dense
and in characteristically different patterns in genera of Onei-
rodidae (for instance Oneirodes, Fig. 1 70A; Dolopichthys. Fig.
170B; Microlophichthys. Fig. 170F; Thaumatichthys, Fig. 169F;
and Cryptopsaras, Fig. 168C, D).
Besides in the completely unpigmented larvae mentioned
above, dorsal pigment is absent in Caulophryne, Neoceratias,
and all Linophrynidae. In all others it is laid down on the antero-
dorsal part of the body. Varying between genera in density and
coverage, it spreads from there and may laterally reach and
overlap the dorsal part of the peritoneal pigmentation and pos-
teriorly the bases of the dorsal and anal fins, in some becoming
confluent with the pigment group of the caudal peduncle.
In occurrence, position and development in relation to larval
length, pigmentation on the caudal peduncle shows very distinct
differences between genera (cf for instance Fig. 1 70) or between
subgenera or species-groups (cf for instance Figs. 167E, F; 169 A,
B; 170C, D).
Other larval structures. — Fe\\\c fins. — In contrast to all other
Ceratioidei, the larvae oi Caulophryne have pelvic fins with well-
developed fin rays (Fig. 167A). The longest of the 3-4 pelvic
fin rays increase in relative length from about 45% SL in the
smallest larvae (3 mm) to about 60% SL in the largest (7.5 mm).
In the only known free-living stage of a metamorphosed male
328
ONTOGENfY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
t 1 I L_
Fig. 167. Ceratioid larvae. (A) Caulophrynidae, Caulophryme sp., sex ?, 6.6 mm; (B) Neoceratiidae. Neoceratias spinifer. sex ?, 6.3 mm; (C-
G) Linophrynidae; (C) Linophnme subgen. Rhizophryme sp., female, 17.5 mm; (D) Haplophryne mollis, metamorphic male, 13.2 mm; (E)
Linophryne subgen. Linophryne sp., male, 3.8 mm; (F) Linophryne subgen. Stephanophryne indica, female 8.6 mm; (G) Borophryne apogon,
male, 4.3 mm lateral and dorsal views. (All from Bertelsen, 1951.)
(7.5 mm) this length is reduced to 28% SL, and pelvic fin rays
are absent in the two known parasitic males (12-16 mm) as well
as in all metamorphosed females (10-109 mm SL).
Illicium, 2nd cephalic ray and caruncles. — In larvae of all fam-
ilies except Caulophrynidae and Neoceratiidae, sexual dimor-
phism in the development of the illicium is present. In females,
the illicium rudiment is club-shaped and protrudes from the
head or from the bottom of a groove in its enveloping skin (Fig.
168A, B); in males it is represented only by the tiny subdermal
rudiment of the illicial bone. Similarly, the external rudiment
of the second cephalic rays of Diceratiidae and Ceratiidae as
well as the caruncles of the latter family are present in the female
larvae and absent in the males (Figs. 168C, D; 169E).
Among the 16 known Caulophryne and the 1 1 known Neo-
ceratias larvae, no sexual dimorphism has been observed (Ber-
telsen, 1951). In Caulophryne, in which metamorphosed fe-
males lack an escal bulb with photophore but have a well-
developed illicium, the rudiment protrudes on the dorsal side
of the head in the same position as in other ceratioid larvae
(Fig. 167A). In Neoceratias. in which the illicium is completely
absent in the metamorphosed females, all larvae have an elon-
gated cylindrical illicium rudiment (pigmented in larger larvae
(Fig. I67B) slightly protruding, in a position unique among
BERTELSEN: CERATIOIDEI
329
Fig. 168. Ceratioid larvae. (A-B) Gigantactinidae. (A) Gigantactis sp., female, 8.5 mm; (B) Rhynchactis leplonema. female, 7.2 mm; (C-E)
Ceratiidae; (C) Cryptopsaras couesi. female, 5.0 mm; (D) Cryplopsaras couesi. male, 5.0 mm; (E) Ceratias holboetli. 7.6 mm; (F) Centrophrynidae,
Centrophryne spinulosa. male, 7.2 mm. (All from Bertelsen. 1951.)
ceratioid larvae, on the tip of the snout, just above the upper
jaw.
Barbels. — In Linophryne. the only ceratioid genus in which the
metamorphosed females have a hyoid barbel, a rudiment of this
is present as an opaque, wart-like thickenmg of the skin in female
larvae of more than about 10 mm SL (Fig. 167C).
The larvae of both sexes of the single known species of the
family Centrophrynidae differ from all other ceratioid larvae in
having a digitiform, hyoid barbel (Fig. 168F). The barbel re-
mains digitiform in the metamorphic males, but after meta-
morphosis it is in both sexes reduced to a low papilla which
gradually disappears in females larger than about 50 mm.
Spines. — Both male and female larvae of the Linophryne sub-
genus Linophryne and the linophrynid genus Borophryne differ
from all other ceratioid larvae in having well-developed, pointed
sphenotic spines (Fig. 167E, G). None of the other spines (pre-
opercular, quadrate, articular, etc.) of the head skeleton char-
acteristic of females of different genera is developed before
larval metamorphosis.
Relationships
Current principal hypotheses.— Thai the Ceratioidei represent a
monophyletic line appears most clearly from the fact that they
all differ from other Lophiiformes in having developed an ex-
treme and unique sexual dimorphism in which the males are
330
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 169. Ceratioid larvae. (A-B) Himantolophidae. (A) Himantolophus groenlandicus gr., female, 6.0 mm; (B) Himanlotophus alhinares gr.,
male, 7.1 mm; (C-D) Melanocetidae; (C) Melanocelus Ijohnsoni. female, 12.0 mm; (D) Metanocetus murrayi. male, 6.0 mm; (E) Diceratiidae,
Diceralias sp., metamorphic female, 10.5 mm; (F) Thaumatichthyidae, Thauinalichlhys sp., female, 6.4 mm. (All from Bertelsen, 1951.)
dwarfed, lack an external illicium, and are furnished with den-
ticular teeth adapted to attach to the female.
We may assume an ogcocephalid or chaunacid-like ancestral
ceratioid which, from the benthic and littoral environment of
its ancestors, has invaded the bathypelagic zone of the ocean.
Probably this evolution has passed through forms in which the
adults were benthic, while the juveniles after metamorphosis
continued the pelagic life of the larvae during adolescence as
for instance found in the family Chaunacidae and as retained
or reestablished in the ceratioid genus Thaumatichthys. This
move to a new adaptive zone has led to a dimorphism which
separates the tasks of the two sexes, the females obtaining ad-
aptations to the bathypelagic conditions of the lophiiform feed-
ing strategy by passive luring, the males being adapted solely to
active search for a sexual partner. In both sexes the change from
benthic to pelagic life has induced a number of changes of which
the most important are: loss of the pelvic fins; a change of the
position and development of their limb-like pectoral fins, now
used for counteracting gravity during swimming rather than for
support and movement on the bottom; and a general trend to
reduce their density by reduction of bony structures and by
retaining the thick subdermal layer of gelatinous tissue present
in the larvae. In the latter character and in the position and
shape of the pectoral fins, they may be regarded as neotenic as
BERTELSEN: CERATIOIDEI
331
_1 L_
Fig. 1 70. Ceratioid larvae. Oneirodidae. (A) Oneirodes sp., female, 8 mm; (B) Dolopichthys sp., male, 5.4 mm; (C) Chaenophryne draco gr.,
female, 4.0 mm; (D) Chaenophryne longiceps. female, 5.5 mm; (E) Pentherkhlhys sp., female, 10.6 mm; (F) Microlophkhthys Tmkrolophus,
female, 9.0 mm. (All from Bertelsen, 1951.)
proposed by Richard Rosenblatt (quoted by Moser, 1981). In
females the changed conditions have led to extreme specializa-
tions of the luring and feeding mechanisms at the expense of
their swimming ability, while in the males this has induced
different specializations in their attachment mechanisms and
sense organs and a development into more streamlined and
efficient swimmers.
The present division of the Ceratioidei into families is based
mainly on revisions by Regan (1912a. 1926) and Regan and
Trewavas (1932). Some changes have been introduced by Ber-
telsen (1951) and Pietsch (1972) resulting in the present rec-
ognition of the 1 1 families listed in Table 89.
The phylogenetic relationships between the families of the
Ceratioidei are still uncertain. The main reason for this is that
most of the derived osteological characters shared by two or
more families are reduction states or loss of parts following the
general trend mentioned above and similarities in such char-
acters may in many cases represent convergent developments.
At the same time most of the diagnostic family characters which
represent new structures or specialization of organs are auta-
pomorphic (and for this reason not included in Table 89). The
more conspicuous of these are: an extreme prolongation of dor-
sal and anal rays of Caulophrynidae; a dense cover of large
papillae on snout and chin of female Himantolophidae; a hyoid
barbel in larvae and juveniles of both sexes of Centrophrynidae;
photophore-bearing. modified dorsal fin rays (caruncles) in fe-
•8
o
c
u
o
O
O
a
2
3
so
BERTELSEN: CERATIOIDEI
333
male Ceratiidae; and very different high specialization of the
illicial and jaw mechanisms of female Neoceratiidae, Thau-
matichthyidae and Gigantactinidae.
The interrelationships of the ceratioid families have been dis-
cussed by Regan (1912a, 1926), Regan and Trewavas (1932),
Bertelsen (1951), and Pietsch (1972,1 979), the latter illustrating
with branching diagrams, alternative proposals for phylogenetic
relationships of the families. However, no detailed analysis or
full discussion of the basis for these proposals has been pre-
sented. For this reason the dendrogram shown in Fig. 1 7 1 should
be regarded only as a very schematic compilation of the ex-
pressed views, following most closely Pietsch ( 1 979: fig. 26) with
some modifications discussed below.
In accordance with Bertelsen (1951) and Pietsch (1979) it is
assumed that sexual parasitism has been established indepen-
dently in different phylogenetic lineages. The observation of a
parasitic male (character 26) in a representative of one of the
1 5 oneirodid genera (Pietsch, 1 976) makes it extremely unlikely
that the five families in which such males have been observed
represent a monophyletic line. (Furthermore, this observation
underlines the possibility that sexual parasitism might be found
in other families as well.) It seems that the evolutionary step
from the temporary attachment of the male to the female, by
means of the denticular teeth present in all ceratioid males (and
resulting in a close and protracted contact between the dermis
of the pair), to a fusion of their tissues is a less drastic event
than it might be supposed and has been established indepen-
dently in different taxa and possibly even facultative in some,
as proposed by Pietsch (1976).
Presence of an escal photophore (no. 1 1) is presumed to be a
synapomorphy separating the other families from Caulophryn-
idae (and ?Neoceratiidae), a primitive sister-group. This implies
that the similarity of some derived character states (nos. 4. 6,
9, 15) to those of one or more of the families Linophrynidae.
Gigantactinidae, and Ceratiidae is due to convergence in these
bone reductions. The alternative, proposed by Pietsch (1979),
that these families were derived from a caulophrynid-like ances-
tor, would imply that the escal photophore has been evolved
independently in two separate lineages. Morphologic and his-
tologic studies of these organs in different families show simi-
larities in such details that this seems extremely unlikely [cf for
instance Brauer. 1908 (Gigantactis); O'Day, 1974 (Oneirodes);
Hansen and Herring, 1977 (Linophryne); and Munk and Ber-
telsen, 1980 (Chaenophryne)].
Based on a number of shared derived character states (nos.
6, 7, 15, 20, and presence of teeth externally on the jaws) it has
been assumed that Neoceratiidae are closely related to Gigan-
tactinidae. However, they differ considerably in other characters
(nos. 5, 9, 13, 14, 21, and 24) and especially in the illicial and
jaw mechanisms of the females. While the complete loss of
illicium in neoceratiids undoubtedly is a derived character state
it remains uncertain whether this family is derived from ances-
tors with or without escal photophores. As discussed in the
following section, some larval characters might indicate the lat-
ter possibility. In reference to this the numerous characters
shared by the two genera of gigantactinids leaves no doubt that
the lack of photophore in Rhynchactis is due to secondary re-
duction (Bertelsen et al., 1981). While none of the highly spe-
cialized families Linophrynidae, Gigantactinidae, and Cerati-
idae appear closely related, their shared derived character states
may indicate a common descendence as shown in Figure 171.
As pointed out by Pietsch (1972) Centrophrynidae has retained
a number of primitive character states but may be more closely
related to Ceratiidae than to any other family, and Thaumati-
chthyidae are most probably derived from an oneirodid-like
ancestor (Bertelsen and Struhsaker, 1977). The remaining four
families Melanocetidae, Himantolophidae, Diceratiidae, and
Oneirodidae appear more similar to each other than the more
specialized families mentioned above, but as their shared char-
acter states are nearly all primitive their interrelationships are
uncertain. The position of Melanocetidae in the dendrogram
(Fig. 171) is based on the presumption that a reduction of the
number of dorsal fin rays to less than 10 is synapomorphic
within the following series of families.
Except for the significance of observed sexual parasitism
the characters of the males have not been considered in previous
discussions of the interrelationships of the ceratioid families.
The presence of denticular teeth shared by all families is a de-
rived character state in relation to all other Lophiiformes. The
absence of such denticles on the snout observed in caulophryn-
ids and neoceratiids may represent a primitive state within the
suborder. In accordance with the classification based on the
characters of the females or shared by the sexes, the males are
highly but very differently specialized in the families Ceratiidae,
Gigantactinidae, and Linophrynidae while the least number of
presumed derived character states are found in Melanocetidae,
Himantolophidae, and Diceratiidae.
Within the families the inter-generic relationships appear close
and relatively simple in the four families divided into two gen-
era. In each of these one of the genera shows more derived
character states in reductions and specializations than the other.
(Rohia in Caulophrynidae; Phrynichthys in Diceratiidae; Tfiau-
matichthys in Thaumatichthyidae; Cryptopsaras in Ceratiidae
and Rhynchactis in Gigantactinidae). Among the five genera of
the well-defined family Linophrynidae, L/«op/;n'««' appears the
most derived (females with photophore carrying barbels). Bor-
ophryne and Acenlrophryne seem closely related to this genus
(very similar osteology and dentition) while each of the genera
Edriolychmts and Photocorynus appear more isolated; the latter
has retained a number of primitive or less derived character
states (nos. 10, 13, 15, 17).
In contrast to the other ceratioid families no conspicuous
distinctive characters have been found which are common to
the large assemblage of genera united in the family Oneirodidae.
However, the presence of quadrate and articular spines in most
of the genera and shared only with the closely related thau-
matichthyids might be significant and their absence in some
genera could be due to secondary reduction. On the basis of
osteological characters the evolutionary relationships of 9 of
the 1 5 genera were studied by Pietsch ( 1 974) and notes on some
of the others have been added by Bertelsen and Pietsch (1975)
and Pietsch (1975). According to these studies Spiniphryne ap-
pears the most primitive of these genera, having retained well-
developed dermal spines, among a number of other primitive
character states. Among the most specialized genera are Lo-
phodolos (reduction or loss of some elements of the skeleton
and enlargements of others) and Chaenophryne (lack of sphen-
otic, quadrate and articular spines, shape of opercular bones and
a unique structure of ossifications; Pietsch, 1975).
Contribution of early life history stages.— Apart from meristic
and osteological characters shared with adults, the larvae of
ceratioid taxa differ from each other only in pigmentation and
to some extent in morphology. As the pigment patterns vary
334
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
greatly within families only the latter characters may be relevant
to the discussion of the relationships of the families.
The assumption by Bertelsen (1951) that the Caulophrynidae
are isolated from all other ceratioids was based mainly on three
larval characters: ( 1 ) presence of pelvic fins; (2) absence of sexual
dimorphism in the illicial rudiments; and (3) lack of a distal
swelling of these rudiments representing the rudiment of an escal
bulb. The two latter character states indicate that the absence
of an escal photophore in caulophrynids is not due to a sec-
ondary reduction. As expressed by Pietsch (1979) who found a
number of additional resemblances between Caulophryne and
less derived Lophiiformes (lophiids and antennarioids): "That
these primitive character states suddenly reappeared in a lineage
that arose from an ancestorderivedinall, is highly improbable."
The possibility, mentioned above, that the neoceratiids may
represent a similar isolated lineage derived from an ancestor
without escal photophore, is based on the same larval characters:
the absence of sexual dimorphism in their illicial rudiments
which lack distal swellings. However, the fact that neoceratiids
and caulophrynids share these primitive character states fur-
nishes no information on their relationship. In other larval char-
acters, especially body shape and size of pectoral fins, the two
families are extremely different.
The assumption that the absence of escal photophore in the
gigantactinid genus Rhynchactis is due to a secondary reduction
is confirmed by the presence of a club-shaped illicial rudiment
in the female larvae.
Little information on the relationships between the ceratioid
families can be obtained from other observed differences in
larval morphology. The greatly enlarged pectoral fins pres-
ent only in gigantactinids and caulophrynids may as assumed
by Pietsch (1979) represent a primitive character state which
has been retained separately in the two lineages. The most con-
spicuous derived character states of the larvae are the extreme
elongation of the body in neoceratiids, the hump-backed shape
of ceratiids, and the barbels of centrophrynids. Being each re-
stricted to a single family they only confirm the distinct sepa-
ration of these lineages.
Within the families, inter-generic comparisons of larvae are
possible only in Gigantactinidae, Linophrynidae, Ceratiidae and
Oneirodidae. Of the remaining seven families, four are mono-
typic and in each of three, which are divided into two genera,
larvae of only one is known. In each of these families very
distinct inter-generic differences in larval pigment patterns have
been found.
This larval character, retained in juveniles of both sexes, has
been one of the main keys to the identification of the free-living
metamorphosed males and thus has contributed considerably
to the concept of the relationships within the ceratioid families.
The fact that separation of larvae (and males) below generic
level has been possible only in those exceptional cases where
intra-generic differences above species rank (subgenera, species-
groups, etc.) have been observed, underlines that within this
suborder the term "genus" indicates a remarkably well-defined
and natural group.
However, little information on phylogenetic relationships
within the families has been obtained from the study of the
larvae. The difficulties in interpreting their character states is
well illustrated in the Linophrynidae. Two apparently typical
derived larval character states occur in this family: (1) well
developed sphenotic spines (within larval Lophiiformes found
only in the linophrynid genus Borophryne and in one of the
three subgenera of Linophryne), and (2) a characteristic sub-
dermal pigment pattern (found only in the linophrynid genus
Haplophryne and in the two subgenera of Linophryne lacking
larval sphenotic spines). If it is assumed very unlikely that these
specializations have evolved independently in different genera
of the same family, the only alternative is that apparently prim-
itive character states are in fact due to three secondary reduc-
tions: (1) lack of sphenotic spines in two L/>!o/)/!n'«e subgenera;
(2) lack of barbels in female Borophryne, making this one more
subgenus oi Linophryne, and (3) lack of subdermal pigment in
one of the subgenera of Linophryne and in Borophryne.
Ceratioids are still very incompletely known and future stud-
ies on additional characters and as yet unknown forms may
bring answers to at least some of the many questions about their
phylogenetic relationships.
Zoological Museum, University of Copenhagen,
Copenhagen 2100 0, Denmark.
Atherinomorpha: Introduction
B. B. COLLETTE
THE superorder Atherinomorpha (Greenwood et al., 1966)
includes the atherinoids (silversides and phallostethids),
cyprinodontoids (killifishes), and beloniforms (halfbeaks and
their relatives), first grouped together by Rosen (1964) as the
order Atheriniformes. The series Atherinomorpha was redefined
by Rosen and Parenti (1981) as including the Atherinoidei (of
uncertain rank), Cyprinodontiformes and Beloniformes.
Utilizing 17 apomorph characters, Rosen and Parenti (1981)
found 1 0 synapomorphies uniting the atherinoids, Cyprinodon-
tiformes, and Beloniformes. Two of these involve early life his-
tory characters: complete separation of embryonic afferent and
efferent circulation by development of the heart in front of,
rather than under, the head and the presence of large demersal
eggs with long adhesive and short filaments and many lipid
globules that coalesce at the vegetal pole. Four additional syn-
apomorphies between the Cyprinodontiformes and Beloni-
formes show the atherinoids to be the plesiomorphic sister group
of these two orders.
Rosen and Parenti (1981) were unable to find derived char-
acters to unite the atherinoids as a monophyletic group but
White et al. (this volume) have discovered two early life history
characters which define the Atheriniformes as the plesiomor-
phous sister group of the Cyprinodontiformes plus Beloni-
formes.
National Marine Fisheries Service Systematics
Laboratory , National Museum of Natural History,
Washington, District of Columbia 20560.
Beloniformes: Development and Relationships
B. B. CoLLETTE, G. E. McGowEN, N. V. Parin and S. Mito
THE Beloniformes (or Synentognathi) is an order of atherino-
morph fishes containing five families. 37 genera, and about
180 species. Species of the Adrianichthyidae inhabit fresh and/
or brackish waters. Most species of the other four families are
epipelagic marine fishes but several genera of Belonidae and
Hemiramphidae are restricted to fresh waters and a few other
genera contain estuarine and freshwater as well as marine species.
Two groups have been recognized under various names by a
series of authors starting with Schlesinger ( 1 909) and continuing
through Regan (1911b), Nichols and Breder (1928), Rosen ( 1 964),
and Collette (1966). Each of these groups contains two families,
the Scomberesocidae and Belonidae in the first, the Hemiram-
phidae and Exocoetidae in the second. Recently, Rosen and
Parenti (1981) expanded the Beloniformes by adding the Ad-
rianichthyidae to the order as a separate suborder Adrianichthy-
oidei, the sister group of the Exocoetoidei (containing two su-
perfamilies Scomberesocoidea and Exocoetoidea).
Development
Eggs
Most beloniform fishes produce large spherical eggs with at-
taching filaments, characters they share with other atherino-
morph fishes (Rosen and Parenti, 1981). Adrianichthyid eggs
are the smallest (1.0-1.5 mm in diameter); followed by exo-
coetids (generally 1.5-2 mm); Hemiramphidae (typically 1.5-
2.5 mm); Scomberesocidae (slightly elliptical, 1 .5-2.5 mm); and
belonid eggs which are generally the largest (most 3-4 mm)
(Table 90). These eggs typically have a homogeneous yolk and
a relatively small perivitelline space. According to Kovalev-
skaya (1982), eggs with long filaments, distributed over the en-
tire sphere of the egg (one filament may be thicker and longer
than the others) should be considered primitive. Such eggs are
found in the Belonidae, some Hemiramphidae, primitive flying-
fishes of the genera Fodialor and Parexocoetus. and also in many
of the highly specialized species of the subfamily Cypselurinae.
Eggs of the Adrianichthyoidei contain numerous small oil
globules which coalesce, at least to some extent, during devel-
opment (Matsui. 1949), as in the Atheriniformes and Cyprino-
dontiformes (Rosen and Parenti, 1981). Exocoetoid eggs either
contain minute, scattered oil globules (Fig. 1 76C) or lack oil
globules (Table 90).
Adrianichthyid eggs have filaments distributed over the entire
chorion, a condition we refer to as uniformly spaced. Most of
these filaments are short, 0.21-0.35 mm in Horaichthys setnai
(Kulkami, 1940), however, on one portion of the chorion they
are as long as or longer than the egg diameter (Fig. 1 72). Pietri
(1983) described these two topographically distinct types of fil-
aments from the chorionic surface of Oryzias latipes as non-
attaching and attaching. Non-attaching filaments showed a reg-
ular distribution over the chorion with an interfilament distance
of about 65-70 /im, and functioned to maintain the integrity of
the egg cluster. Attaching filaments were located at one pole of
the egg forming a clump of about 25 filaments that united with
those of neighboring eggs to anchor the egg cluster to the gon-
oduct of the female. In Oryzias melastigma, the attaching fil-
aments also anchor the eggs to the female (Job, 1940) or to
filamentous algae.
The eggs of most scomberesocids {Scomberesox, Namchthys
and Elassichthys) are pelagic, without long filaments. Eggs of
Scomberesox (Fig. 1 73A), however, have short bristles that ap-
parently represent remnants of chorionic filaments (see Boehlert,
Table 90. Eggs of Beloniformes Fishes. Much of this information is based entirely on illustrations from the cited references.
Filaments
Diameter
(mm)
Oil
globule
Taxon
Arrangement
Number
Length (mm)
Remarks and sources
Adrianichthyidae
Horaichthys setnai
1
Numerous co-
2 types:
Demersal, Kulkar-
alesce into 1 5-
uniform
Type 1. most nu-
0.21-0.35
ni, 1940
30
localized
merous
Type 2, many
>egg dia.
Oryzias javanicus
Numerous co-
alesce into 1
2 types
Ahlstrom notes
O. latipes
1.27-1.3
Numerous co-
2 types:
Ahlstrom notes;
alesce into 1
uniform
localized
Type 1 , nonat-
taching. most
numerous
Type 2, attach-
mg, -25
Matsui, 1949;
Pietri, 1983;
Hart et al.,
1984
O. luzonensis
-1.5
Numerous co-
alesce into 1
2 types
Ahlstrom notes
O melastigma
1.0-1.2
30-40 coalesce
2 types:
Job, 1940
into 1
uniform
localized
Type 1, most nu-
merous
Type 2, many
<egg dia.
>egg dia.
335
336
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 90. Continued.
Filaments
Diameter
Oil
Taxon
(mm)
globule
Arrangement
Number
Length (mm)
Remarks and sources
Scomberesocidae
Cototabis saira
1.7 X 1.9 (off
round)
No
Polar cluster
lateral
12-15
1
Demersal. SWFC,
unpublished and
original
C. adocetus
-2.5
None
Orton, 1964
Scomberesox
Off round.
No
Uniform or
Numerous
Short, rigid
Pelagic, Hardy,
saurus
range of
mean diame-
ters =
2.32 X 2.52
groups
1978a
S. simulans
None
Hubbs and Wis-
ner, 1980
Belonidae
Ablennes hians
3.0-3.5; mean
No
Uniformly-
1-6 per tuft; to-
>egg dia.
Hardy, 1978 and
= 3.16 ripe
spaced tufts
tal 37-59 (N =
original
ovarian
6)
Belone belone
3.0-3.5
No
Uniform
60-80
4-18
Demersal, Russell,
1976
Pseudolylosurus
1.2-1.4; mean
CoUette, 1974a
microps
= 1.23
and original
Slrongylura exilis
2.3-2.8; mean
= 2.5 (run-
ning ripe)
No
Uniform
18-30
Longer than egg
dia.
Original
S. Jluvialilis
2.9-3.2; mean
= 3.1 (ovari-
an)
Original
S. hubbsi
2.50-3.14;
mean = 2.75
(ovarian)
Collette, 1974c
S. incisa
3.5-4.6; mean
= 3.9 (ovari-
an)
Original
S. krefftii
2.7-4.0; mean
= 3.4 (ovari-
an)
Original
S. marina
3.5-3.6
No
Uniform
Numerous
Variable, but
generally <egg
diameter
Demersal, Hardy,
1978a
S. notata
3.67-4.18;
mean = 3.95
Uniform
Demersal. Breder,
1959
S. slrongylura
2.5
No, but one or
Uniform
Numerous 57 in
All long, but 1-2
Demersal, Job and
more clear
illus.
areas longest
Jones, 1938
vesicles project
into yolk
Tylosurus acus
3.22-4.0
Uniformly-
spaced tufts
2-3 per tuft
Longer than egg
dia.
Demersal, Hardy,
1978a
T. a. melanotus
3.2-3.5
No
Uniformly-
spaced tufts
2-3 per tuft; to-
tal @ 100
2-3 X egg dia.
Mito, 1958
T. crocodilus
4.0-4. 1
Minute
Numerous
Long
Demersal, Masu-
rekar, 1967
T. punctulalus
3.5-4.3; mean
= 3.9
Original
Xenenlodon canci-
2.9-3.2
No
Foster, 1973 and
la
original
Hemiramphidae
Dermogenvs pusil-
Viviparous
No
Mohr, 1936a;
lus
Brembach, 1976
Euleptorhamphus
1.1 (ovarian)
"Disorderly"
6.0-6.5
Demersal; Parin
viridis
and Gorbunova,
1964
Hemira mph us fa r
2.8-3.1
Present
Breder and Ro-
sen, 1966
He. marginatus =
Mean = 2.59
No
8-12 tufts
4-6/tuft
Both > & < egg
Talwar, 1968
He. lutkei
diameter
COLLETTE ET AL.: BELONIFORMES
337
Table 90. Continued.
nio mAtdr
Oil
Filamenls
Taxon
(mm)
globule
Arrangement
Number
Length (mm)
Remarks and sources
Hemirhamphodon
Viviparous
Mohr, 1936c;
Brembach, 1976
Hyporhamphus
2.2-2.6
14-16 tufts
3-4/tuft
Sudarsan. 1968b
guoyi
Hy. capensis
-1.6
Present
Smith, 1933b
Hy. intermedius
1.3-1.4
Minute
Bipolar
3^/tuft,
1 polar
7
>25
Uchida et al.,
1958
Hy. sajori
2.1-2.3
Minute
Bipolar
4-6/tuft.
1 polar
20-60
Uchida et al.,
1958
Hy. unifascialus
-2.0
Several
>egg diameter
Semibuoyant;
Hardy, 1978a
Hy. xanthoplerus
1.12-1.44;
mean = 1.23
(ovarian)
Collette, 1982b
Melapedalion
breve
Nomorhamphus
Viviparous
Mohr, 1936b;
Brembach, 1976
Oxyporhamphus
1.9-2.3
Uniform
Many
0.05-0.06
Pelagic; Kovalev-
conve.xus
skaya, 1965
O. micropterus
1.8-2.1
No
Uniform
74-120
0.08-0.12
Pelagic; Imai,
Rhychorhamphus
georgii
R. malabaricus
-1.5
Zenarchopterus
Not viviparous
Z. roberlsi
3.0-3.5; mean
= 3.25 (ovari-
an)
Exocoetidae
Cheilopogon (A.)
1.4-1.6
agoo
Ch. (Ch.)pinnali-
1.57-1.70;
barbalus califor-
mean = 1.64
nicus
Ch. (Ch.)pinnali-
1.9-2.1
barbalus japoni-
cus
Ch. (Proc.) cyanop-
1.2 (ovarian.
Icrus
prob. not ful-
ly ripe)
Ch. (Proc.) katop-
tron
Ch. (Proc.) ni-
2.0-2.2
gricans
1.8 (maximum
ovarian)
Ch. (Proc.) spilop-
1.79-2.17; No
terus
mean = 2.02
Ch. (Pi.) heterurus
1.86 No
Ch. (Pt.)h. doeder-
2.0-2.2
leini (type 1 )
Ch. (Pl.)h. doeder-
1.8-2.2
leini (type 2)
Ch. (Pi.) hiikeni
-1.6
Ch. (Pi.) melanu-
1.8-1.9
rus
Bipolar ?
0.07-0.08
1959
Kovalevskaya,
1965
Demersal; Koval-
evskaya, 1965
Demersal; Deva-
nesen, 1937
Mohr, 1926;
Brembach, 1976
Collette, 1982c
Uniform
34-66
Demersal, Imai,
1960
Uniform
-60
Hubbs and Kam-
pa, 1946
Uniform
56-78
-10
Demersal, Imai,
1959
Uniform
Gibbs and Staiger,
1970
Uniform
Numerous
0.5-0.55
Pelagic, Kovalev-
skaya, 1965
Uniform
Numerous
<egg diameter
Pelagic, Pann and
Gorbunova,
1964
Gibbs and Staiger,
1970
Uniform
12-19
Long
Demersal, Vijay-
(usually 13)
araghavan, 1975
Uniform
Numerous
Long
Demersal, Hardv,
1978a
Bipolar
30-86
10
Demersal, Imai,
1959
Uniform
30-48
8-11
Demersal, Imai,
1959
45-1-
Hubbs and Kam-
pa, 1946
Uniform
Gibbs and Staiger,
1970
338
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 90. Continued.
Diameter
(mm)
Oil
globule
Arrangemeni
Length (mm)
Remarks and sources
Ch. (Pt.) unicolor 1.5-1.6
Cypselurus (Cy)
1.8-2.1
naresii
Cy. (Cy.) opisthops
1.5-1.6
hiraii
Cy. (Poec.) cy-
1.2 (ovarian.
anopterus
prob. not ful-
ly ripe)
Cy. (Poec.) starksi
1.6
Exocoetus mono-
2.8-3.3 (ovari-
cirrhus
an)
E. obtusirostris
2.8-2.9 (ovari-
an)
E. volitans
1.7-2.0
2.7-3.0
Fodiator acutus
1.4-1.7; mean
= 1.53 (ovari-
an)
Hirundichthvs (D.)
1.4-1.5
rondeleti
H. (//.) affinis
1.6 + 0.1 (ripe
ovarian)
H. (//.) coroman-
Mean= 1.87
delensis
H. (//.) oxycepha-
1.5-1.7
lus
H. (//.) speculiger
1.53-1.87;
mean = 1.65
2.05-2.15
(ovarian; after
swelling)
Parexocoetus bra-
1.2-1.5 (ovari-
chypterus bra-
an)
chyplerus
P. memo memo
1.7-1.8
Prognichthys brevi-
pinms
No
Minute
Minute
Bipolar
Uniform
Uniform
Uniform
Uniform
None
None
None
None
Uniform
Bipolar
Bipolar
Bipolar
Bipolar
Bipolar
Bipolar
Bipolar
Uniform
Uniform
10-12
18-20
40-52
55-88
40-52
9 (from illustra-
tion)
90-100
1, many
?
8-14
mean = 1 1
1,3-5
5-12
9-18
10-25
18-20
30
7-10
1
6-10
10-18
20
Short
Long
15-21
>egg diameter
Short
40-1-
102.5, 4.6
1.1
4
0.15
5
0.17-0.18
< egg diameter
35-40
1.5-2.0
1.5-2.0
3.5-8.0
Short, less than
egg diameter
Demersal, Gorbu-
nova and Parin,
1963; Pann and
Gorbunova,
1964
Demersal, Imai,
1959
Demersal, Imai,
1959
Gibbs and Staiger,
1970
Demersal, Imai,
1959
Pelagic, Kovalev-
skaya, 1964
Pelagic. Breder,
1938
Pelagic, Bruun,
1935
Kovalevskaya,
1964
Demersal, Breder,
1938
Demersal, D'An-
cona. 1929
Kovalevskaya,
1972
Demersal, Evans,
1962
Demersal,
Vijayaraghavan,
1973
Demersal, Imai,
1960
Demersal, Munro,
1954
Parin and Gorbu-
nova, 1964
Demersal, Imai,
1959
Demersal, Tsuka-
hara and Shiok-
awa, 1957
Kovalevskaya,
1982
this volume). Cololabis is the only scomberesocid with fila-
ments, a polar cluster of relatively long filaments plus a single
long lateral filament (Fig. 173B). Cololabis eggs typically are
attached to floating objects such as kelp.
Belonid eggs have filaments (Table 90). typically long, nu-
merous and uniformly spaced over the chorion (Fig. 1 74A). In
at least one species, Strongylura strongylura. some filaments are
markedly longer than others (Fig. 1 74B), as in the adrianich-
Ihyids. The filaments on eggs of Tylosaurus acus are arranged
in uniformly distributed tufts containing 2-3 filaments each (Fig.
174C).
Hemiramphids have eggs with attaching filaments (Rhyn-
chorhamphus. Hyporhamphus, and Hemiramphus. Fig. 175A),
pelagic eggs with very short uniformly-spaced filaments (O.xy-
porhamphus. Fig. 175B), or are viviparous (Hemirhampliodon,
Dermogenys, Nomorhamphus). Filaments were not reported on
ovarian eggs of Dermogenys by Flegler (1977) and we did not
note their presence in Hcmirhamphodon or Zenarchopterus but
this needs to be checked more thoroughly. Kovalevskaya (1965)
reported that filaments on Rhynchorhamphus georgii have a
bipolar arrangement; however, this is not clear in her illustration
(Fig. 175 A). Talwar (1968) and Sudarsan (1968b) reported the
COLLETTE ET AL.: BELONIFORMES
339
Fig. 172. Adrianichthyidae egg. Horaichthys selnai. (From: Kul-
kami, 1940.)
filaments of what they called Hemiramphus marginatus (Fig.
175C) and Hyporhamphus qitoyi, respectively, to be grouped in
tufts. The filaments in a tuft may be of different lengths (e.g..
He. marginatus).
Most exocoetids have eggs that are attached with thin thread-
like filaments to objects floating in the water column or to sea-
weed growing near shore. The size and structure of the eggs and
the size, nature, and location of the filaments vary among species.
The eggs of Fodiator and Prognichthys have uniformly-spaced
filaments (Fig. 176B). Filaments on Hirundichthys eggs have a
bipolar arrangement. One species, H. coromandelensis, has three
types of filaments (Vijayaraghavan, 1973), but they are grouped
in a bipolar distribution. This type of egg has a single long (103
mm), stout filament arising from the basal pole, which is sur-
rounded by 3-5 medium length (.v = 4.6 mm) filaments. Five
to 1 2 short (.V =1.1 mm) filaments are located at the distal pole
(Fig. 1 76C). Chorionic filaments in Parexocoetus. Cheilopogon
and Cypselurus vary. Some species have uniformly-spaced fil-
aments, whereas others have a bipolar arrangement with the
filaments usually longer at one pole than at the other. Unlike
all other flying fishes, species of Exocoetus have eggs with a
smooth membrane, devoid of filaments (Fig. 176A).
Larvae
A relatively long incubation period is typical of the Beloni-
formes (Kovalevskaya, 1982). The eggs develop for one to two
weeks, and the larvae are well formed and capable of actively
capturing food at hatching. Time of development is comparable
in pelagic and attaching eggs. Following a pattern similar to that
Fig. 173. Scomberesocidae eggs. (A) Scomberesox saurns. SWFC
Cr. Est 1-4 Sta. Surf. I; (B) Cololabis saira. SWFC CalCOH. (Original.)
reported for egg size, belonids hatch at the largest sizes (6.8-
14.4 mm) followed by hemiramphids (4.8-11 mm), scomber-
esocids (at least as small as 6.0-8.5 mm), exocoetids (3.5-6.1
mm), and adrianichthyids (3.5-4.5 mm).
Gut length differs between the two suborders. Adrianichthy-
oid larvae have a short gut, as in Atheriniformes and Cyprin-
odontiformes, 40-50% of standard length (Fig. 177A). Exocoe-
toid larvae are generally elongate and have a straight gut extending
approximately two-thirds the standard length (Fig. I77B-E, G,
and H).
Presence of a preanal finfold appears to be plesiomorphic.
Job's (1940) illustration of a yolk-sac adrianichthyoid, Oryzias
melastigma. shows a preanal finfold (Fig. 1 77A), but Kulkami's
( 1 940) illustration of a yolk-sac Horaichthys setnai does not. A
preanal finfold is present until after formation of all fins in the
belonids, hemiramphids and scomberesocids (Fig. I77B-E). The
situation in the Exocoetidae is not clear. Most published illus-
trations of exocoetids do not show a preanal finfold. Ones that
B
Fig. 174. Belonidae eggs. (A) Slrongylura exilis. LACM 43475-1;
(B) Slrongylura slrongylura; (C) Tylosurus acus. (From: A. Original. B.
Job and Jones, 1938. C. Mito. 1958.)
Fig. 175. Hemiramphidae eggs. (A) Rhynchorhamphus georgii; (B)
Oxyporhamphus muroplerus micropterus; (C) Hemiramphus margin-
alus. (From: A. Kovalevskaya, 1965. B. Imai. 1959. C. Taiwan 1968.)
COLLETTE ET AL.: BELONIFORMES
341
do (Evans, 1962— Hirundichthys affinis; Vijayarghavan, 1973 —
H. coromandelensis; Vijayaraghavan, 1975 — Cypselurus spilop-
terus and Kovalevskaya, 1965 — Cheilopogon katoptron) are of
embryos or newly hatched larvae and, except for Kovalevskaya
(1965), were hatched in the laboratory (Fig. 177H). In these
examples the preanal finfold was small and soon lost. We have
examined field-collected yolk-sac Cheilopogon (pTes\imab\y Ch.
pinnatibarbatus californicus) without finding a preanal finfold
(Fig. 177G). Perhaps some exocoetids have a preanal finfold,
but lose it soon after hatching. If so, most field-collected spec-
imens may have already lost the preanal finfold by the minimum
sizes typically illustrated.
Fin formation generally begins during the embryonic stages
or soon after hatching. In fact, flexion of the caudal fin precedes
hatching in flyingfishes (Ahlstrom and Moser, 1980). In the
scomberesocids, belonids and hemiramphids, caudal, dorsal and
anal fins generally form first followed by the pectorals and lastly
the pelvics. Pectoral and pelvic buds as well as dorsal and anal
anlagen are typically present at hatching in exocoetids. Pectoral
fins form last in exocoetids, after the pelvic fins.
Belonids, scomberesocids and exocoetids generally hatch with
heavy, uniform pigmentation formed or forming over essentially
the entire body (Fig. 1 778, C, and G). Exceptions are the fresh-
water needlefish Xenentodon cancila. which has 9-10 saddle-
shaped dorsal aggregations plus a ventrolateral stripe (Fig. 1 77D)
and some exocoetids of the genera Parexocoetus and Cheilo-
pogon. which have patterns somewhat reminiscent of the hem-
iramphids (compare Fig. 177E and H). This pattern consists of
three rows of melanophores on each side of the body, one dorsal,
one lateral and one ventral. Two hemiramphids reported to be
exceptions to this are Hyporhamphus quoyi and Hemiramphus
marginatus. These species hatch with pigment over the entire
body; a pattern reminiscent of most other beloniforms. The
pigment pattern in adrianichthyids resembles that in hemiram-
phids except dorsally where the adrianichthyids have a single
middorsal row of melanophores (Fig. 181 A), similar to the con-
dition observed in Atheriniformes (see White et al., this volume)
rather than the double row typical of most hemiramphids (Fig.
177F).
Specialized Ontogenetic Stages
During post-embryonic development, beloniform fishes
undergo a number of complex changes.Their larvae differ fairly
strongly from juveniles and the juveniles are frequently unlike
adults. Juveniles of related species frequently differ more from
each other than do larvae or adults. In this section, notable
ontogenetic changes are described for several character suites
in the four families of the Exocoetoidei. Adrianichthyoids lack
specialized ontogenetic stages.
Jaws, beaks, and barbels
Scomberesocidae.— Juveniles (20-40 mm SL) have slightly
elongate upper and lower jaws but no prominent beaks (Fig.
18 IB; Hubbs and Wisner, 1980: fig. a). At about 60 mm
SL, both upper and lower jaws, but especially the lower jaw,
elongate in Scomberesox and Namchthys. Elongation continues
in both taxa to 100-120 mm SL. Both jaws elongate almost
equally in Scomberesox; the lower jaw exceeds the upper in
Fig. 176. Exocoetidae eggs. (A) Exocoetus volilansAM) t'odiator acu-
lus pacificus: (C) Hirundichthys coromandelensis. (From: A. Parin and
Gorbunova, 1964. B. Breder, 1938. C. Vijayaraghavan, 1973.)
342
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Fig. 177. Beloniform larvae. (A) Adrianichthyidae: Oryzias melastigma. 4.3 mm; (B) Scomberesocidae: Cololabis saira. SWFC 5009-50. 1 10,
5.1 mm SL; (C-D) Belonidae; (C) Strongylura exilis. LACM 42756-5, 8.6 mm SL; (D) Xenemodon cancila. 9.6 mm SL ANSP 124230: (E-F)
Hemiramphidae: (E-F) Hyporhamphus rosae. 5.7 mm SL, LACM 42870-5; (G-H) Exocoetidae: (G) Cheilopogon pinnatibarbatus califonucus.
LACM IP77-3, 3.7 mm SL; (H) Cheilopogon katoptron. 3.2 mm SL. From: (A) Job, 1940; (B-G) Original: (H) Kovalevskaya, 1965.
Nanichthys. A slight beak develops in Cololabis; Elassichthys
does not develop a distinct beak.
Belonidae. — Most species of Belonidae pass through a "half-
beak" stage in which the lower jaw, but not the upper jaw, is
greatly elongate. Juveniles of Belone belone remain in the half-
beak stage for a longer time than other needlefishes. This has
led directly to four synonyms of Belone belone described as
halfbeaks (CoUette and Farin, 1 970: 1 6-17). Plotting the relative
length of the lower jaw extension, as a percentage of head length
against body length (Fig. 1 78), shows that lower jaw extension
in B. belone may be nearly 150% of head length at 25 mm BL
(body length) and decreases to less than 10% by 175 mm BL.
Petalichthys and Platybelone (Fig. 179E) also remain in the
halfbeak stage for a long time. The duration of the halfbeak
stage varies among species of Strongylura (Fig. 1 79C and F).
Comparative development of Platybelone {as Strongylura long-
leyi), Strongylura marina, S. nolata, and two species of Tyto-
surus, (T. acus and T. crocodilus. Fig. 179G and J) was illus-
trated by Breder ( 1 932: figs. 7 and 1 0, plates 1 and 2). Tylosurus
crocodilus (Fig. 1 79J) completely lacks a halfbeak stage, upper
and lower jaws growing at the same rate from larval to adult
stages of development (Breder, 1932: plate 2, fig. 2, as T. ra-
phidoma). The South American freshwater genus Belonion
(maximum size 42 mm body length) is characterized by ma-
turing while still in the halfbeak stage (Fig. 1 79A-B) and was
considered paedomorphic by Collette (1966).
Hemiramphidae.— Adults of four genera of halfbeaks lack the
elongate lower jaw that characterizes most members of the fam-
ily. The lower jaw extends only 1.5-1 1.0 mm beyond the upper
jaw throughout the size range in Arrhamphus (Collette, 1974b).
COLLETTE ET AL.: BELONIFORMES
343
g
CO
z
UJ
t-
X
UJ
<
— )
a:
I
»-
O
z
UJ
_j
Q
<
LU
I
150
100
50
-..^^^^
T~^^^s#^-^^
• i .•% •
••
25
100
200
300
400
500
BODY LENGTH (mm)
Fig. 178. Relative growth of upper jaw in Belone belone. Lower jaw extension as a percent of head length plotted against body length. Inset
is of a 43.7 mm BL B. belone from Ireland in the "halfbeak" stage. (From: Collette and Parin. 1970.)
The lower jaw is even shorter in Melapedalion and virtually
absent in adult Chriodorus and Oxyporhamphus. Chriodorus
looks superficially more like an atherinid than a halfbeak, hence
its specific name of at herinoides. Adult Oxyporhamphus resem-
ble flyingfishes because of the enlarged pectoral fins. Juveniles
of all four genera have a distinct beak. Arrhamphus, Melape-
dalion. and Chriodorus have always been considered halfbeaks.
Oxyporhamphus has usually been considered an exocoetid or
placed in a separate family.' Even with its short beak, Arrham-
phus varies geographically in beak length: Arrhamphus s. scler-
olepis of northern Australia has a proportionately shorter lower
jaw (up to 20 times in head length) than does A. sclerolepis
krefflii of southern Queensland and New South Wales (up to 1 1
times in head length, see Collette, 1974b: fig. 4).
Exocoetidae.— The two most primitive genera of flyingfishes,
Fodiator and Parexocoetus, have an elongate lower jaw (Parin,
' Parin (1961), although still recognizing the Oxyporhamphidae as
valid, clearly showed that Oxyporhamphus is a halfbeak, even though
it has a straight margin to the upper jaw instead of triangular as in other
halfbeaks. Two developmental characters support placement of O.vr-
porhamphus in the Hemiramphidae: a preanal fin fold is present in
larvae (absent or lost soon after hatching in Exocoetidae) and the pelvic
fins form last (pectoral fins form last in Exocoetidae).
1961; Kovalevskaya, 1982). This clearly is a beak m juvenile
(15-55 mm SL) Fodiator, which like several genera of halfbeaks,
lose their beaks as they grow larger (Fig. 1 8 1 C and Breder, 1938:
figs. 5 and 6E). A beak is present in Parexocoetus mento (Imai,
1959). Small ( 1 9-20 mm) P. hrachypterus have a pair of barbels
that are attached to the ventral surface of the beak and obscure
it (Fig. 182). Thus, a beak which is absent in advanced flying-
fishes, is present in both primitive genera.
Juvenile stages of many exocoetids develop barbels on the
lower jaw (Table 91). Barbels range from relatively short to
longer than body length (Fig. 181D-I). Parexocoetus mento
does not develop a barbel nor do species of Prognichthys and
Hirundichthys (Kovalevskaya, 1982). Paired barbels develop in
Parexocoetus hrachypterus and in all species of Cheilopogon
(Fig. 181D, G-I; Kovalevskaya, 1982). In species of Cheilo-
pogon (subgenus Procypselurus, Ch. nigricans group), the bar-
bels consist of a thick strand with a leathery fold branching oflT
it in the form of a lobe (Parin, 1961; Kovalevskaya, 1982). In
small specimens of Ch. cyanopterus the barbel may be complex
and have 2-3 flaps. Members of Cheilopogon (subgenus Mac-
ulocoetus) have flattened barbels, joined together at the base.
These may be large. The barbels of C/!«7opo^o« (subgenus Pten-
ichthys) range from short (in Ch. heterurus doederleini) to long
(in Ch. unicolor). The barbels in Cheilopogon pinnatibarbatus
344
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
T^#^"^
.m^
._^ ^ I 1, II I ijii,.,i.iiijjui:(xc. ■"
-'^-^-■:'^^^^^
<. . ^S%l
COLLETTE ET AL.: BELONIFORMES
345
(subgenus Cheilopogon s. str.) are flaplike and fringed (Fig. 1811).
Kovalevskaya (1982) considered this to result from the fusion
of paired barbels and our examination of Ch. pmnatibarbatus
calijorniciis supports this. The barbel is single in Cypselurus
(subgenus Cypselurus s. str.) and Exocoetus monocirrhus (Fig.
1 8 1 F; Parin, 1961; Kovalevskaya, 1982). Some species of Cyp-
selurus (subgenus Poecilocypselurus, Cy. poecilopterus and Cy.
starksi) do not develop a barbel, nor do the remaining two
species of Exocoetus, E. obtusirostris and E. volilans.
Melanistic dorsal fin lobe
Pelagic members of three families (all except Scomberesoci-
dae) develop prominent melanistic lobes in the dorsal fin. The
lobe is in the posterior part of the dorsal fin in the Belonidae
and Hemiramphidae but in the middle of the fin in the Exo-
coetidae so presence of the lobe is not necessarily homologous.
Beionidae.— A blennes and Tylosurus are characterized by hav-
ing a prominent enlarged melanistic lobe in the posterior part
of the dorsal fin (Fig. 179D, G-J). Other genera of needlefishes
(Fig. 1 79) lack any trace of this posterior dorsal lobe. Breder
(1932: plates 3-5) illustrated the development of this posterior
lobe in T. acus and T. crocodilus and its absence in Strongylura
and Platybelone. Parin (1967) left an Australian species difficult
to place in either Tylosurus or Strongylura in a monotypic genus
described by Whitley, Lhotskia gavialoides. A juvenile with a
well-developed posterior dorsal lobe, captured by Collette. con-
vinces us that it is a species of Tylosurus (Fig. 1 79H). The lobe
is apparently sloughed off in Tylosurus crocodilus (Breder and
Rasquin, 1952), resorbedin T. aoM (Breder and Rasquin. 1954),
and retained in adult Ablennes.
Hemiramphidae.— Juveniles of Hemiramphus and Oxypo-
rhamphus develop a darkened posterior lobe on the dorsal fin
(Fig. 180) similar to that present in two genera of needlefishes,
Ablennes and Tylosurus.
Exocoetidae. — In juveniles of many species oi Cheilopogon. the
middle portion of the dorsal fin develops a melanistic lobe (Fig.
1 8 1 H). This is reminiscent of the adult stage of Parexocoetus
and Eodiator acutus.
Body bars
Juveniles of some species in three exocoetoid families (all
except Scomberesocidae) have vertical bars on their body.
Belonidae.— Juveniles of two species of Tylosurus, T. gavi-
aloides (Fig. 179H) and T. acus (see Collette and Parin. 1970:
fig. 12) and .Ablennes hians have bars. These bars are retamed
in adult .-iblennes as is the posterior dorsal fin lobe.
Hemiramphidae. — The 10 species of the genus Hemiramphus
all have a series of broad vertical bars on the body (Fig. 180A-
E) at some stage of their development. Body bars are retained
for different periods of time during development: all body bars
are lost before 105 mm SL in He. lutkei and He. depauperatus
(Parin et al., 1980: fig. 32), before 120 mm SL in He. bermu-
densis and He. brasiliensis (Collette, 1962: fig. 1), but are re-
tained past 175 mm SL in He. balao; one blotch is retained
throughout life in He. robustus, and all are retained in He. far.
Pelvic fin pigment
All 10 species of Hemiramphus also have pigmented pelvic
fins as juveniles (Fig. 183). The patterns of pelvic fin pigmen-
tation divide the genus into two species groups, one with pig-
mentation concentrated proximally on the fin (balao group. Fig.
183, top two rows), the other with pigment absent basally and
concentrated distally (J'ar-brasiliensis group. Fig. 183. bottom
row). Body bars and pelvic fin pigmentation are absent in Hy-
porhamphus.
Exocoetidae. — In late larval and juvenile stages of many flying-
fishes, Exocoetus, Cheilopogon (at least some species in all sub-
genera except possibly Paracypselurus, for which we lack data),
Cypselurus (subgenus Poecilocypselurus— see Imai, 1959), and
Hirundichthys oxycephalus (Imai, 1960) transverse stripes de-
velop on the abdomen and sides of the body which disappear
(sometimes leaving traces) in adults. The coloration of the larvae
and particularly of the juveniles of flyingfishes is diverse, and,
as a rule, differs greatly from the coloration of adults. A partic-
ularly bright variegated coloration is characteristic of young of
neritic species living among algae (Parin, 1961; Kovalevskaya,
1982).
Relationships
Beloniformes
The Beloniformes were defined by 7 characters by Rosen and
Parenti (1981:16). Meristic characters for the beloniform genera
are summarized in Table 92. A cladogram for the families and
higher taxa of the Beloniformes is presented as Fig. 184.
Adrianichthyoidei
Rosen and Parenti (1981) defined the adrianichthyoids by 5
characters. Larval adrianichthyids also differ from exocoetoids
in having a shorter preanal distance, 40-50% of standard length.
Rosen and Parenti (1981) included the Horaichthyidae and Ory-
ziidae in the Adrianichthyidae. By this definition the Adrianich-
thyidae includes four genera, .Adrianichthys, Horaichthys, Ory-
:ias and Xenopoecihis with a total of 1 1 species (Nelson, 1976).
These fishes inhabit fresh and/or brackish waters from India
and Japan to the Indo-Australian Archipelago.
Fig. 179. Halfbeak stages of Belonidae, arranged by relative length of upper Jaw. (A) Belonion apodion Collelle: USNM 199540; Brazil, Borba;
29.4 mm BL; (B) Belonion dihranchodon Collette; USNM 199463; Venezuela, Rio Atabapo; 38.2 mm BL; (C) Strongylura marina (Walbaum),
USNM 189006; Nicaragua; 23.5 mm BL; (D) Ablennes hians (Valenciennes); USNM 188843; Gulf of Honduras; 36.1 mm BL; (E) Platybelone
argatus argalus (Le Sueur) USNM 198102; 39°28'N, 69°30'W; 96 mm BL; (F) Strongylura p.v;fa (Girard); SIO H47-158-23A; Calif, La Jolla;
72.5 mm BL; (G) Tylosurus acus acus (Ucepede); USNM 1 98402; 38°00'N, 65''25'W; 1 30 mm BL; (H) Tylosurus gavialoides (Castelnau); USNM
226666; Australia, New South Wales; 72.5 mm BL; (1) T. choram (Forsskal); USNM 147438; Red Sea; 95.0 mm BL; (J) T. c. crocodilus (Peron
and Le Sueur); USNM 198407; 37°08'N, 66°14'W; 96.3 mm BL. A-G, 1-J drawn by Mildred H. Carrington; H by Keiko Hiratsuka Moore; A-
C from Collette (1966: tig. 1).
346
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 180. Juvenile banded stages of five species of Hemiramphus and Oxyporhamphus micropterus. (A) Hemiramphus balao (Le Sueur);
USNM 200592; off Cape Hatteras, North Carolina; 53.7 mm SL; (B) He saltator (GWben and Starks); SIO 55-247; Gulf of Panama; 49.5 mm
SL; (C) He. depauperatus Lay and Bennett; Hawaii Inst. Mar. Biol.; Hawaiian Is.; 41.4 mm SL; (D) He. far (ForsskSl); USNM 148020; Persian
Gulf; 47.0 mm SL; (E) He. hrasiliensis (Linneaus); USNM 188748; off North Carolina; 50.0 mm SL; (F) Oxyporhamphus micropterus similis
Bruun; USNM 159032; Gulf of Mexico; 41.2 mm SL. A-B, D-F drawn by Mildred H. Camngton; C by Keiko Hiratsuka Moore.
COLLETTE ET AL.: BELONIFORMES
347
Table 91. Larvae of Beloniform Fishes. Much of this information is based entirely on illustrations from the cited references.
Hatching
length (mm)
Adrianichthyidae
Horaichthys selnai
Oryzias melasligma
Scomberesocidae
Cololabis saira
C. adocetus
Scomheresox saurus
S. simulans
Belonidae
Abennes hians
Belone belone
Platybelone argalus
Strongylura manna
S. strongylura
Tylosurus acus
T. crocodilus
Xenentodon cancila
Hemiramphidae
Dermogenys pusillus
Hemiramphus hrasiliensis
He. marginatus
Hemirhamphodon pogonognathus
Hyporhamphus guoyi
Hy. limbatus
Hy. intermedius lutkei?
Hy. sajori
Oxyporhamphus convexus
O. microplerus microplerus
Exocoetidae
Cheilopogon (A.) agoo
Ch. (Ch.) pmnalibarbalus
californicus
Ch. (Ch.) pinnatibarbatus
japonicus
Ch. (M.) spilonotopleriLs
None
3.5-4.0
4.0-4.5
Slight or none
None
6.0-8.5
Present by 15-17 mm
Present by 40 mm
Present bv 13.9 mm
9.0
Present bv 1 8 mm
Present by 47 mm
9.2-14.4
Present by 14.3 mm
6.75
Present by fifth
day
10.16
Present by 14.1 mm
10.7-12.0
Present by 15.2 mm
10.5
Present
-11
5-7 SL
Present by 13.0 mm
5.85
-11
-6.3
Present at hatching
Present by 1 2.0 mm
4.8
Present by 10.7 mm
Present by 12.3 mm
Present by 14.7 mm
7.7
Present by 6-8 mm
Ch. (A/.) spiloplcrus
Ch. (M.) sulloni
Ch. (Proc.) cyanoplenis
Ch. (Proc.) exsiliens
Ch. (Proc.) katoptron
Ch. (Proc.) nigricans
Ch. (Pt.)furcalus
Ch. (Pt.) helcrunis
Ch. (Pt.) h. doederleini
Ch. (Pt.) umcolor
Cypselurus (Cy.) comatus
Cy. (Cy.) naresii
4.5-5.3
None
4.1-4.8
None
:an = 4.45
4.6-5.8
None
None
4.52
None
None
4.5?
5.2-6.1
None
None
None
None
None
None
None
None
None
None
None
None
None
None
Pair, short, present by
19.0 mm
Fan-like, complex with 14
fimbriae
Fan-like with flaps, present
by 20.1 mm
Pair, very long, present
by 10.4 mm
Pair, flattened and
joined at the base
Pair, complex on smaller
individuals, then long
filaments
Pair, medium, present by
14.5 mm
Short, present by 10.2 mm
Pair, long, complex
Pair, medium length, develop
on individuals between
7.7 and 18 mm
Pair, present by 19 mm TL
Pair, short, present by
18.1 mm
Pair, long, present by
5.8 mm
Single, medium length
Single tape-like, very long
with appendages at the
base
Kulkami, 1940
Job, 1940
Hubbs and Wisner, 1980 and
original
Hubbs and Wisner, 1980
Hardy, 1978a; Fahay, 1983
Hubbs and Wisner, 1980
Mito, 1966
Russell, 1976
Original
Hardy, 1978a
Job and Jones, 1938
Hardy, 1978a
Masurekar, 1968
Foster, 1973
Soong, 1968
Hardy, 1978a; Berkeley
and Houde, 1978
Talwar, 1968
Soong, 1968
Sudarsan, 1968b
Nair, 1952b; Job and
Jones, 1938
Uchidaet al., 1958
Uchidaet al., 1958
Kovalevskaya, 1965
Chrapkova-Kovalevskaya,
1963; Kovalevskaya,
1965c
Imai, 1960
Hubbs and Kampa, 1946
Imai, 1959
Imai, 1959
Vijayaraghavan, 1975
Kovalevskaya, 1982
Breder, 1938; original
Imai, 1959
Kovalevskaya, 1965
Kovalevskaya, 1982
Hildebrand and Cable,
1930
Hardy, 1978a
Imai, 1959
Gorbunova and Parin,
1963
Breder, 1938
Imai, 1959
348
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 91. Continued.
Halching
length (mm)
Cy. (Cy.) opisthops hiraii
Cy. (Poec.) poecilopterus
Cy. {Poec.) slarksi
Exocoetus monocirrhus
E. ohlusirostris
E. volitans
Fodiator aculus
Hirundichthys (D.)
albtmaculatus
H. (D.) marginatus
H. (D) rondeteti
H. (//.) affinis
H. (H.) coromandelensis
H. (//.) o.xycephalus
H. (H.) speculiger
Parexocoelus brachyplenis
brachyplerus
P. mcnto memo
Prognichlhys gibbifrons
P. sealei
4.5-5.1
None
Smgle, petalous
by 17.5 mm
present
Imai, 1959
None
Absent
Imai, 1959
5.2
None
Absent
Imai, 1959
None
Single, long
dev
elops on
Kovalevskaya,
1964; Imai.
individuals between 16.0
1959
and 18.6
•nm
None
Absent
Kovalevskaya,
1964
None
Absent
Kovalevskaya,
1964
Present
by
14.6
mm
Absent
Original
None
Absent
Kovalevskaya,
1972
None
Absent
Kovalevskaya,
1972
None
Absent
Imai, 1960
None
Absent
Evans, 1962; Breder, 1938
3.47-4.23
None
Absent
Vijayaraghavan, 1973;
Kovalevskaya, 1972
None
Absent
Imai, 1960
None
Absent
Imai, 1960
Present by
18.1
mm
Pair, short
Imai, 1959
4.5-5.2
Present by
23.8
mm
Absent
Tsukahara and Shiokawa,
1957; Imai,
1959
None
Absent
Origmal
None
Absent
Imai, 1960
Exocoetoidei
Defined by six characters by Rosen and Parenti (1981: 16).
We here add two developmental characters: oil droplets in egg
minute or absent and preanal distance of larvae increased to
about 66% of standard length.
The Exocoetoidei is undoubtedly a monophyletic group.
However, vainous opinions have been expressed as to which
group within the suborder is the most primitive. A number of
authors have considered the Hemiramphidae to be the most
generalized family based largely on the fact that needlefishes
and primitive flying fishes {Fodiator) pass through an ontoge-
netic halfbeak stage during development. Parin (1961) and Ro-
sen ( 1 964) supported this viewpoint. On the other hand, Nichols
and Breder (1928) and Breder (1932) considered the family
Belonidae the most primitive. To resolve the directionality of
the "halfbeak" stage (suite four), three additional character suites,
each suite consisting of several correlated transformation series,
were considered. Apomorphic character states are numbered
higher than plesiomorphic states on Fig. 184.
The first suite involves pharyngeal tooth plate fusion, trans-
formation series A-B. State Al is close opposition of left and
right fifth ceratobranchial tooth plates characteristic of more
primitive Atherinomorpha and the Adrianichthyoidei. State A2
is the fusion of left and right lower pharyngeal bones into a tooth
plate in the Exocoetoidei. Series B state 1 is when the third
upper pharyngeals are separated by a gap. State B2 is when they
are joined but not fused in the Exocoetoidea. State B3 is the
complete fusion of the third upper pharyngeals into a tooth plate
in the Hemiramphidae.
The second suite involves loss of gill arch skeleton bones,
transformation series C-D. State CI is presence of the fourth
epibranchial, C2 its loss in the Beloniformes. State Dl is the
presence of the fourth upper pharyngeal tooth plates, D2 their
loss in the superfamily Exocoetoidea.
The third suite involves reduction in the cephalic lateralis
system, transformation series E-F (data from Parin and Astak-
hov, 1982). The cephalic system is more complete in the Scom-
beresocoidea than in the Exocoetoidea, including the presence
of a premaxillary canal (EI), an autapomorphy unique among
teleosts. The pre-, supra-, and post-orbital system is continuous
across the top of the head in state 1 . There are short interruptions
in the system in state 2 in the Belonidae. The postorbital section
is lost in state 3 and secondary bony canals are lost in state 4,
both charactenstic of the superfamily Exocoetoidea.
We now return to the fourth suite of transformation series
and resolve the directionality of the "halfbeak" stage. The fourth
suite includes elongation of upper and lower jaws and presence
of barbels in juveniles, transformation series G-I. State Gl is
Fig. 181. Late larval and early juvenile stages of beloniform fishes. (A) Adrianichthyidae: Oryzias melastigma, 12 mm; (B) Scomberesocidae:
Cololabis adocetus. SWFC 7205 J-20.145, 25 mm SL; (C-I) Exocoetidae: (C) Fodiator acutus pacificus. SWFC FB-62-242, 15.4 mm SL; (D)
Cheitopogon unicolor. 18.5 mm SL; (E) Cypselurus comatus. 25 mm SL; (F) Exocoetus monocirrhus. SWFC FB-62-203, 27 mm SL; (G) Pare.xocoetus
brachyplerus, 43.5 mm SL; (H) Cheitopogon cyanopterus. 54 mm SL; (I) Cheitopogon pinnatibarbatus japomcus. 80 mm SL. From: (A) Job, 1940;
(B, C, and F) Original; (D) Gorbunova and Parin, 1963; (E, G and H) Breder, 1938; (F) Kovalevskaya, 1964; and (I) Abe, 1954.
COLLETTE ET AL.: BELONIFORMES
349
350
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
■aeur'
/iecir-
Fig. 182. Exocoetidae, Parexocoetusbrachypterusbrachypterus. (A)
18.1 mm; (B) 19.7 mm.
lower jaw elongate in juveniles and adults, G2 elongate only in
juveniles, and G3 never elongate, even in juveniles. Presence
of an elongate lower jaw is considered a synapomorphy of the
suborder Exocoetoidei because it is present in the most gener-
alized members of each of the four families. This is supported
ontogenetically by its presence in juveniles and loss in adults of
four genera of Hemiramphidae and in the two least derived
subfamilies of Exocoetidae.
Transformation series H involves elongation of the upper jaw.
It is most parsimonious to hypothesize the elongation of the
upper jaw as a synapomorphy (H2) of the superfamily Scom-
beresocoidea. Thus, the absence of an elongate upper jaw is
plesiomorphous (HI) in the Exocoetoidea.
Transformation series I is the development of barbels in ju-
veniles of advanced flyingfishes. State 1 1 is the absence of bar-
bels. If we consider barbels in flyingfishes to be derived from
the pair of cutaneous lappets on the lower jaw of needlefishes,
halfbeaks, and primitive flyingfishes, the most generalized state
of this character is the presence of two separate barbels, 12 (Fig.
185). This supposition is supported ontogenetically by two ju-
venile Parexocoetus brachypterus brachypterus. The smaller one
(Fig. 182A, 18.1 mm) has a short beak from the ventral surface
of which a pair of small barbels develop in the larger one (Fig.
182B, 19.7 mm). Fusion into a single barbel (13) and secondary
loss of the barbels (14) are more derived states. Loss of the
barbels has apparently occurred independently in the three most
advanced subfamilies of the Exocoetidae.
Scomberesocoidea
The superfamily is defined by two derived characters: pres-
ence of a premaxillary canal, unique among teleosts; and upper
jaw at least slightly elongate. Other diagnostic characters in-
clude: third pair of upper pharyngeal bones separate, fourth
upper pharyngeal bone usually present, scales on body small.
The Scomberesocoidea differ from the Exocoetoidea in four
characters of the acoustico-lateralis system (Parin and Astak-
hov, 1982). The cephalic system is more complete in the Scom-
beresocoidea than in the Exocoetoidea.
Scomberesocidae
Defined by one derived character: dorsal and anal fins fol-
lowed by a series of finlets. Other diagnostic characters include:
upper and lower jaws only slightly elongate, teeth small; pectoral
branch of lateral line absent, posttemporal simple. There are
four more differences in the acoustico-lateralis system between
the Scomberesocidae and the Belonidae (Parin and Astakhov,
1982).
Four monotypic genera were recognized by Hubbs and Wisner
(1980): Scomhereso.x and its dwarf derivative Nantchthys. and
Cololabis and its dwarf derivative Elassichthys. All sauries are
marine holoepipelagic fishes. Scomberesox reaches the largest
size, 450 mm SL, Nanichthys reaches 126 mm; Cololabis reaches
350-400 mm, Elassichthys only 68 mm. The two dwarf taxa
differ convergently from Scomberesox and Cololabis in losing
one ovary and the swimbladder and in having fewer vertebrae,
branchiostegal rays, pectoral fin rays, and gill rakers. Rather
than recognizing four monotypic genera, we recognize two evo-
lutionary lines in the family by considering Nanichthys as a
synonym of Scomberesox and Elassichthys a synonym of Col-
olabis as previously suggested by Parin (1968).
Belonidae
Defined by one derived reductive character: interruptions in
the cephalic lateralis system. Other diagnostic characters in-
clude: no finlets following dorsal and anal fins; both upper and
lower jaws usually elongate and studded with relatively large
sharp teeth; pectoral branch of lateral line present; posttemporal
forked.
The Belonidae contain 10 genera and 32 species (Collette,
1966, 1974a, 1982a). Four genera are monotypic: the southern
African Petalichthys. the worldwide Ablennes and Platybelone.
and the Asian freshwater Xenentodon. Belone contains two east-
em Atlantic species. Three genera are restricted to freshwaters
of South America: Pseudotylosiirus (two species), Potamorrha-
phis (three), and Belonion (two). Tylosurus contains five species
of strictly marine species; Slrongylura 14 species, some marine,
some estuarine, and three strictly freshwater.
The genera Belone and Petalichthys appear to be most gen-
COLLETTE ET AL.: BELONIFORMES
351
<^^
352
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 92.
Number of Dorsal, Anal, and Pectoral Fin Rays, Vertebrae and Gill Rakers on the First Gill Arch in the Genera of
Beloniformes.
No- ofspp.
D
A
p,
Venebrae
Family and genus
Precaud,
Caud-
Total
gr
Adrianichthyidae
Adnanichlhys
1
17
25
16
Horaichthys
1
6-7
22-32
Oryzias
7
6-9
17-25
11-15
29-32
13
Xenopoecilus
2
11-13
21-27
11-13
Belonidae
Belone
2
16-20
19-23
11-14
48-54
25-30
75-84
27-52
Petatichthys
1
16-19
21-23
10-12
46-47
26-27
72-74
27-35
Platybelone
1
11-17
15-21
10-13
39-48
23-29
62-76
7-14
Tylosurus
5
18-27
17-25
11-15
41-65
23-33
67-96
0
Ablennes
1
22-26
24-29
11-15
51-63
30-37
82-97
0
Strong\'liira
14
12-23
12-27
9-13
34-57
19-34
53-90
0
Xenentodon
1
14-19
14-19
10-12
35-40
21-25
57-62
0
Pseudotylosurus
2
13-16
14-19
8-11
42-47
25-28
67-73
0
Potamorrhaphis
3
27-43
24-39
6-8
35-42
28-44
64-85
0
Belonion
2
11-14
12-15
5-6
32-37
19-23
52-59
0
Scomberesocidae
Scomberesox
1
15-18
17-21
12-15
39-43
24-28
64-70
34-51
Nanichthys
1
14-16
17-20
10-11
35-38
22-26
58-62
19-26
Cololabis
1
14-18
18-21
12-15
37-40
24-29
62-69
32-42
Elassichthys
1
14-18
16-21
8-11
32-35
21-24
54-59
15-21
Hemiramphidae
Arrhamphus
1
13-16
14-17
12-14
28-32
16-19
18-25
45-50
Chriodonts
1
15-18
15-17
12-14
31-33
18-19
49-51
20-24
Melapedalion
1
15-17
14-16
12-13
33-35
17-18
51-52
25-31
Hemiraniphus
10
11-15
9-14
10-13
30-41
16-19
50-59
25-48
Rhyruiwrhamphiis
4
13-17
12-16
10-12
37-40
16-19
54-59
47-78
Hyporhamphus (Hyporhamphus)
23
12-17
13-17
10-13
28-37
16-20
45-55
20-53
Hyporhamphiis (Rcporhamphus)
II
13-18
13-19
10-13
31-42
15-20
49-61
26-47
Oxyporhamphus
2
12-15
13-17
11-13
30-33
17-19
47-50
26-36
Euleplorhamphus
2
20-25
20-25
7-9
44-46
26-29
70-75
24-35
Zenarchoplerus
17
10-16
8-14
7-11
25-36
11-18
38-51
11-18
Dennogenys
3
8-12
14-17
9-13
21-24
16-17
38-40
11-14
Hemirhamphodon
3
14-23
8-9
8-9
24-26
14-16
38-41
11-16
Nomorhamphus
2
12-13
13-17
11-13
21-24
17-19
40-42
0
Exocoetidae
Fodiator
1
9-11
10-12
13-14
25-26
14-16
39-41
29-33
Parexocoetus
2
9-13
10-14
12-14
21-25
14-16
36-40
23-32
Exocoetus
3
13-15
12-15
14-17
24-27
16-20
42-44
23-35
Cypselurus
11
10-14
7-10
13-17
28-20
13-15
39-48
17-24
Cheilopogon
18
9-16
7-12
13-17
25-30
12-16
41-51
19-28
Prognichlhys
4
10-13
8-10
14-19
26-34
12-17
43-45
21-28
Hirundichlhys
8
9-13
9-13
15-20
26-31
14-19
42-47
23-32
eralized (CoUette and Berry, 1965; Parin, 1967), having well-
developed gill rakers, large scales, comparatively weak canine
teeth and other generalized characters. Belone also is charac-
terized by the most completely developed cranial lateralis sys-
tem (Parin and Astakhov, 1982). Of the other genera, the great-
est number of primitive characters are found in Platybelone.
characterized at the same time by several specialized features
(in particular, the well-developed cutaneous lateral keel on the
caudal peduncle). The remaining three genera of marine needle-
fishes (Strongylura, Ablennes. and Tylosurus) are more ad-
vanced but their relationships have been interpreted differently
by Collette and Berry (1965), Parin (1967), and Astakhov ( 1 980).
The freshwater genera of needlefishes, in the opinion of all au-
thors, have been derived from the genus Strongylura or its
ancestor and are secondary freshwater fishes.
Exocoetoidea
The superfamily is defined by one derived character and three
losses: third pair of upper pharyngeal bones united into a plate;
fourth upper pharyngeal bone lost; postorbital section and sec-
ondary bony canals of cephalic lateralis system lost. Other di-
agnostic characters include: scales on body large, lower jaw fre-
quently elongate but upper jaw never significantly elongate; and
premaxillary canal absent.
COLLETTE ET AL.: BELONIFORMES
353
BELONIFORMES
C2
Fig. 184. Cladogram of the Beloniformes. See text for explanation
of character transformation series A-H.
o
(b
0)
.c
<b
o
(3
■^
(3
.c
.C
(Li
0
o
o
•V
.C
^^
o
o
o
o
o
(Z)
■^
^°
<^*
'^
Exocoetidae
Fig. 185. Cladogram of the Exocoetidae. I. Swimbladder extends
into haemal canal. 2. Pectoral fins enlarged. 3. Lower jaw not elongate
in adults. 4. Loss of preanal finfold. 5. Barbels present in juveniles (12).
6. Pectoral lateral line branch lost. 7. Beak lost in juvenile (G3). 8.
Pectoral fins greatly enlarged. 9. Swimbladder extends far into haemal
canal. 10. Pelvic fins enlarged. 1 1. Egg filaments lost.
Hemiramphidae
Defined by one derived character: third pair of upper pha-
ryngeal bones ankylosed into a plate. Other diagnostic characters
include: pectoral fins short or moderately long; premaxillae
pointed anteriorly, forming a triangular upper jaw (except in
Oxyporhamphiis); lower jaw elongate in juveniles of all genera,
adults of most genera; parapophyses forked; swimbladder not
extending into haemal canal.
The Hemiramphidae contains 12 genera and at least 80 species
(Parin et al., 1980). Four genera, the first three monotypic (Ar-
rhamphus, Chriodorus. Melapedalion, and Oxyporhamphus)
have very short or no beaks. Euleplorhamphus and Oxypo-
rhamphus contain two offshore species each. Zenarchoptenis,
Dermogenys, Hemirhamphodon. and Nomorhamphus contain
about 25 sexually dimorphic Indo-West Pacific estuarine or
freshwater species. Three of these genera (Dermogenys. Hemi-
ramphodon. and Nomorhamphus) are viviparous and have the
anal fin of the male modified into what Brembach (1976) has
termed an andropodium. Hemiramphus (with 10 species) is a
world wide manne genus. Rhynchorhamphus (with 4 species)
has fimbriate nasal papillae and is confined to Indo-West Pacific
marine waters. Hyporhamphus. the most speciose genus, in-
cludes two subgenera, Hyporhamphus with 23 species, Repo-
rhamphus with 1 1 species. Some of these are marine, some es-
tuarine, and some freshwater. All genera are characterized by
particular lateral line characters (Parin and Astakhov, 1982).
Exocoetidae
Defined by one derived morphological character and three
derived early life history characters: swimbladder extending into
haemal canal; lower jaw of adults not elongate; preanal finfold
reduced or lost; and pectoral fins form last. Other diagnostic
characters include: third pair of upper pharyngeal bones coales-
cent, the plate readily separating into its left and right compo-
nents; pectoral fins long; premaxillae with straight anterior mar-
gin; parapophyses simple, not forked.
The family Exocoetidae contains 7 genera and about 50-55
species (Parin, 1961) which have been placed in four subfamilies
(Bruun, 1935; Parin, 1961; Fig. 185): Fodiatorinae containing
354
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
only Fodiator acutus (with two subspecies; reaching 195 mm
SL); Parexocoetinae with two species of Parexocoetus (reaching
140 mm SL); Exocoetinae with three species of Exocoetus
(reaching 200 mm SL); and Cypselurinae with four genera—
Prognichthys (4 species; reaching 1 90 mm SL), Cypselurus sensu
stricto (11 species; reaching 260 mm SL), Cheilopogon (not
differentiated from Cypselurus by some authors; 1 8 species; con-
tains the largest species of flyingfishes, some reaching 380 mm
SL), and Hinmdichthys (8 species; reaching 190 mm SL; in-
cludes the more specialized subgenus Danichthys which was
recognized as a genus by Bruun and others). All are strictly
marine, mostly in tropical and subtropical waters.
Similarities in the skeletal structure (Parin, 1961) and lateralis
system (Parin and Astakhov, 1982) between Exocoetus and the
Cypselurinae (Cheilopogon, Cypselurus, Prognichthys, and Hi-
rundichthyus) indicate that differentiation of Exocoetus from
the main stem took place significantly later than separation of
the primitive short-winged flyingfishes (Fodiator and Parexo-
coetus). There is particular interest in the interrelationships within
the subfamily Cypselurinae. One problem concerns whether
Cypselurus should be accepted in the wide sense (Bruun, 1935;
Staiger, 1965; Gibbs and Staiger, 1970) or divided into two
genera, Cypselurus and Cheilopogon (Parin, 1961). The diag-
nostic differences between these two genera are not simple.
Therefore, Parin herein presents the following definition: "lower
jaw usually a little shorter than the upper; at least some jaw
teeth tricuspid; juveniles with a single chin barbel or without
barbels" in Cypselurus, and "lower jaw a little longer than upper,
teeth mostly unicuspid or with smaller supplementary cusps
laterally; juveniles with two barbels which may be fused into a
napkin-like appendage" in Cheilopogon. Each genus contains
groups of species, several of which were distinguished by Bruun
(1935) or Parin (1961) at the level of subgenera.
The similarities and differences between species groups are
most noticeable in the juvenile stages and form the basis of the
systematics of the Cypselurinae worked out by Parin (1961). If
we consider barbels in flyingfishes to be derived from the pair
of cutaneous lappets on the lower jaw of needlefishes, halfbeaks,
and primitive flyingfishes, the most generalized state of this
character is the presence of two separate barbels. Their deriv-
atives are fusion into a single appendage or complete loss. In
the speciose genus Cheilopogon, according to the classification
of Parin (1961), the juvenile stages of most intrageneric group-
ings—the subgenera Procypselurus (composed of the Ch. ni-
gricans and Ch. cyanoptenis groups), Maculocoetus, and Abe-
ichthys— are characterized by a pair of barbels, sometimes joined
at their bases, and presence of an enlarged melanistic dorsal fin
("Parexocoetus stage"). In juveniles of the subgenus Cheilopo-
gon. the dorsal fin is greatly enlarged, but the barbels are fused
into a fringed appendage. In the subgenus Ptemchthys, paired
barbels remain but the " Pare.xocoetus stage" is lost (present only
in Ch. longibarbits, which, apparently should be removed from
this subgenus). The subgenus Paracypselurus is somewhat in-
termediate between Cheilopogon and Cypselurus. Juveniles have
paired barbels and an enlarged dorsal fin, but adults are closer
to Cypselurus \n structure ofthejaw and other characters (except
absence of tricuspid teeth).
Summary
There is a considerable amount of information available on
the early life stages of beloniform fishes. Specialized structures
such as egg filaments, barbels, beaks, and melanistic dorsal fin
lobes have systematic value. It is pleisiomorphous for the eggs
of beloniform fishes to have chorionic filaments (Rosen and
Parenti, 1981). One or more loss events presumably gave rise
to the apomorphous condition, an absence of chorionic fila-
ments, seen in the dwarf sauries (Cololabis adocetus and Scom-
heresox simidans) and in the flyingfishes of the genus Exocoetus.
The development of a beak during some life stage is a derived
feature that occurs in all belonids, scomberesocids (except C
adocetus) and hemiramphids, and the two most primitive ex-
ocoetid genera (Fodiator and Pare.xocoetus). It is never found
in the adrianichthyids. Presence of a beak is a synapomorphy
for the Exocoetoidei and supports Rosen and Parenti's (1981)
division of the Beloniformes into two suborders, the Adrianich-
thyoidei (no beak) and the Exocoetoidei (beak). A second char-
acter that supports this is relative length of the gut at hatching,
40-50% standard length in Adrianichthyoidei and approxi-
mately 66% in the Exocoetoidei. The superfamily Scombere-
socoidea differs from the Exocoetoidea in having a premaxillary
lateral line canal and in having the upper jaw at least slightly
elongate.
(B.B.C.) National Marine Fisheries Service Systematics
Laboratory, National Mliseum of Natural History,
Washington, District of Columbia 20560; (G.E.M.)
Section of Ichthyology. Los Angeles County Museiim
OF Natural History, 900 E.xposition Boulevard, Los
Angeles, California 90007; (N.V.P.) P.P. Shirshov
Institute of Oceanology, Academy of Sciences of the
U.S.S.R., Krasikova Street 22, Moscow 1 1 72 1 8, U.S.S.R.;
(S.M.) Research Division, Fisheries Agency, Ministry
OF Agriculture, Forestry and Fisheries, Government
OF Japan, 2-1, 1-Chome, Kasumigasekj, Chiyoda-Ku,
Tokyo, Japan.
Atheriniformes: Development and Relationships
B. N. White, R. J. Lavenberg and G. E. McGowen
IN the latest statement on the evolutionary relationships of
the atherinomorph fishes (Rosen and Parenti, 1981), mono-
phyly could not be established fijr the Atherinoidei. No derived
characters could be offered to unite the constituent families
(Atherinidae. Bedotiidae, Isonidae, Melanotaeniidae, Phallo-
stethidae, and Telmatherinidae) and the group term Atheri-
noidei was dropped in favor of a listing convention placing them
in Division I of a general classification of the series Atherino-
morpha. In this report, two synapomorphic character states are
described that suggest that the Division I fishes are indeed a
monophyletic group and the group name Atheriniformes is res-
urrected for this assemblage. This new order is defined by a
derived larval pigmentation pattern and a reduction in preanal
length that persists from hatching through early flexion. Except
for this modification, the classification and familial designations
of Rosen and Parenti (1981) are accepted here.
Development
Eggs
Information on atheriniform egg morphology is assembled in
Table 93. The smallest atheriniform egg known, that of Atherion
elymus. measures 0.55-0.58 mm in diameter (Nakamura, 1936).
The largest eggs average approximately 2.3 mm in diameter and
are found in the genus Aihehna (Marion, 1 894a; Kanidev, 1961).
Numerous oil globules are found in the yolk of most species.
Usually, the globules aggregate at the vegetal pole and may
coalesce into a single droplet that comes to lie near the heart.
In Bedotia geayi. the globules form an equatorial ring two hours
after fertilization and reach the vegetal pole by the blastula stage
(N. R. Foster, Fish. Wildl. Serv., Michigan, pers. comm.). At
fertilization, there may be as few as one oil globule, in Chiros-
toma bartoni (de Buen, 1 940), or as many as 115, in Leuresthes
tenuis (David. 1939).
Although absent in Leuresthes, Atherion, and Bedotia. cho-
rionic filaments are found on the eggs of most species. The eggs
can be bound together in a mass by these filaments or attached
singly to a substratum. There is only one filament on the eggs
of Eurystole eriarcha. Menidia extensa. and Telinathenna la-
digesi but most species have more. The filaments can be scat-
tered over the surface of the egg, as in Atherinops and Atheri-
nopsis. or gathered together in a tuft as in .-itherina. Membras,
Odontesthes. Melanotaenia. Memdia menidia and Afenidia ber-
yllina. In Menidia beryllina, one filament is much enlarged;
being longer and thicker than the others making up the tuft
(Hildebrand, 1922). Until more information is available, it will
be difficult to assess the phylogenetic significance of this vari-
ation in the size, number and placement of the chorionic fila-
ments. No pattern is readily apparent. In some cases, not all of
the species assigned to a genus have the chorionic filaments
arranged in the same way. In both Menidia and .Austroinenidia
there are species in which the filaments are collected in a tuft
and species in which they are randomly scattered. Two egg types
may occur in .tt/urinops affinis. There are approximately 6 fil-
aments attached at one end to the chorion (Crabtree, pers. comm.)
(Fig. 186A) or 40-78 looped filaments attached by both ends
to the egg surface (Curless, 1979). This unusual occurrence of
two egg types in Atherinops may support the contention that
there is more than one species in the genus (Hubbs, 1918).
The remarkable ovarian egg of Eurystole eriarcha is unlike
that known for any other atheriniform species. It averages 1.7
mm in diameter and is pigmented, with a brownish band swirl-
ing over its surface (Fig. I86B). Arising from the pigmented
portion of the chorion are numerous small anchor-shaped ped-
icels. Each egg has one major filament arising from the side of
one of these unusually shaped pedicels (Fig. 187 upper). Some
eggs appear to have a small number of finer filaments similarly
attached to some of the other pedicels, but the majority of these
chorionic projections do not have attached filaments. Each fil-
ament can become entangled in the pedicels of its own and
neighboring eggs (Fig. 1 86B). The pedicels and small depressions
that serve as bases of attachment are unpigmented.
The vitelline circulatory system of all atheriniform species
examined is simple, unbranched and looping. This pattern is
common within the Atherinomorpha. However, the vitelline
circulatory system of the cyprinodontoids is characterized by a
complex branching pattern.
Larvae
Morphologically, the larvae of the atheriniform fishes are much
less variable than the eggs. Development is direct and the known
larvae are similar in appearance (Fig. 188). Pectoral fin buds
appear in embryos. Throughout the Atheriniformes the preanal
finfold regresses as the origin of the dorsal finfold comes to be
more posteriorly placed. After hatching, fin rays develop in the
caudal fin ventral to the upturned tip of the vertebral column.
Next, the pectoral, anal and second dorsal fins become rayed
and then the pelvic fin buds develop. Finally, spines appear in
the first dorsal and anal fins. The gut is short; with the preanal
length averaging one-third the body length (NL or SL) from
hatching through the time of flexion. In all atheriniform larvae
known, except Odontesthes debueni. preanal length is less than
40% of body length at flexion. Preanal length in Odontesthes
debueni is 45% of body length (Fig. 188 A). All known ather-
iniform larvae are similarly pigmented. Melanophores occur on
the top of the head and dorsally and laterally on the gut. Typ-
ically, a single row of melanophores occurs mid-laterally along
the body, as well as on the dorsal and ventral margins.
Within the Atheriniformes, the total number of vertebrae
ranges between 2 1 and 60. with the typical number of precaudal
vertebrae being 22-23 (Ahlstrom notes; Rosen and Parenti,
1981). Meristic data are compiled for 89 atheriniform species
and subspecies in Table 94.
Information is available on the early life history of a variety
of atheriniform species. The larvae ofAlherinomorus insularum
(Miller etal., 1979), Iso hawaiiensis (MxWcr tlaX., 1979), Odon-
testhes regia (Fischer, 1 963) and Menidia menidia (Hildebrand,
1922) follow the normal mode of atheriniform development
355
356
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 93. Egg Characteristics of the Atheriniformes.
Diameter
Vitelline
Taxon
(mm)
Oil globules
circulation
Filaments
Sources
Atherinidae
Atherina boyeri
1.72-1.76
Coalesce and move
towards heart
Unbranched,
looping
—
Sparta, 1942b
Athenna hepselus
2.0-2.5
Coalesce and move
Unbranched,
Tuft
Marion, 1894a; Breder
towards heart
looping
and Rosen, 1966
Alheruia mochon pontica
2.3
Coalesce and move
towards heart
Unbranched.
looping
—
Kanidev, 1961
Atherina presbyter
1.85-1.95
—
—
Numerous
Miller, 1961
Alhennops affinis
Channel Islands and Mainland
1.62
40-78 attached
at both ends
to the chorion
Curiess, 1979
Mainland California
~6 attached at
one end to the
chorion
Crabtree, pers. comm.
Atherinopsis californtensis
2.0-2.5
Numerous
-12
Clark, 1929
Atherion elymus
0.55-0.58
Several large (0.16 mm)
and small
Absent
Nakamura, 1936
A ustromenidia incisa
1.7-2.15
Numerous, small
Present,
scattered
de Ciechomski, 1972
Austromenidia regia
1.8-2.15
> 50. coalesce into 1
Unbranched,
looping
Tuft of 5-10
Fischer, 1963
Chirosloma bartoni
1.0-1.1
1; or several
coalesce into 1
Unbranched,
looping
Tuft
de Buen, 1 940
Eurystole enarcha
1 attached to one
of numerous
anchors
This study
Leuresthes tenuis
1.5-1.6
25-115 that coalesce
Unbranched,
looping
Absent
David, 1939
Membras vagrans
0.8-1.1
1 2 that coalesce
Unbranched,
looping
Tuft
Kuntz, 1916
Menidia beryllina
0.75
7-10 that coalesce
Unbranched,
looping
Tuft
Hildebrand, 1922
Menidia extensa
0.6
—
—
1
Davis and Louder, 1969
Menidia menidia
1.25
7-10 that coalesce
Unbranched,
looping
Tuft
Hildebrand, 1922
Menidia notata
2.1
Several of differing
—
Tuft of 4
Ryder, 1883
Odontesthes bonariensis
-1.0
size
Several that coalesce
-10
Minoprio, 1944
Odontesthes debueni
1.65-1.86
10-20 globules that
coalesce
Unbranched,
looping
Tuft of 6-9
Fischer, 1963
Telmatherinidae
Telmatherina ladigesi
—
—
-
1
Breder and Rosen, 1966
Bedotiidae
Bedolia geayi
-2.0
Numerous
Unbranched,
Absent
Neal R. Foster,
looping
Phallostethidae
Gulaphallus mirabilis
Gulaphallus falcifer
Melanotaeniidae
Melanotaenia maccuUochi
Pseudomugil signata
Pseudomugil signifer
1.0-1.1
1
1.5
1.6
70-80 do not coalesce
10-15 large; numerous
small, do not
coalesce
Numerous, cluster but
do not coalesce
Numerous
8 scattered
Many in a tuft
Tuft
Unbranched,
looping
pers. comm.
Villadolid and
Manacop, 1934
Manacop, 1936
Neal R. Foster,
pers. comm.
Neal R. Foster,
pers. comm.
Breder and Rosen, 1 966
WHITE ET AL.: ATHERINIFORMES
357
with some minor exceptions. The larva of /. hawaiiensis (Fig.
188B) has a deeper body than any other known atheriniform
larva and A. msidarum (Fig. 188C) lacks the ventral melano-
phore series typical of the order. The ventral melanophore series
is also absent in Odontesthes regia (Fischer, 1963) which has
only sparse midlateral pigment at hatching.
In Menidia. dorsal pigmentation can be sparse or even lacking
(Hildebrand, 1 922). In M. memdia. it has been reported that a
double row of dorsal melanophores occurs in older larvae (ca.
11 mm) (Lippson and Moran, 1974). However, in a smaller
flexion specimen (8 mm), we found a double row of melano-
phores in the area of the dorsal fin, but only a single row anterior
to the fin. The dorsal melanophore row is interrupted by the
dorsal fin in other atheriniform larvae as well. This pattern also
occurs in the melanotaeniid genus, Psendomugtl (Foster, pers.
comm.). It is not unusual for single melanophores to be divided
by a developing fin in other fishes and it is assumed here that
the more complex distribution of dorsal pigment in Menidia
and Pseudomugil is a variation on the simpler pattern seen in
Atherinomorus, Iso, Odontesthes and most other atheriniform
larvae. In Melanotaenia, a single dorsal row develops. The larval
morphology of Melanotaenia and Pseudomugil closely resem-
bles that of the other atheriniform fishes (Foster, pers. comm.).
Larval Dentatherina mercen differ from all other known ath-
eriniform larvae but resemble larval Oryzias in having a double
row of melanophores on the nape. The melanophores on the
dorsal surface of the trunk are unpaired except where they are
interrupted by the developing dorsal fins. The larva of Bedolia
geayi (Fig. I88D) has the single dorsal melanophore row and
short gut typical of the Atheriniformes. Interestingly, the ventral
pigment series of Bedotia is paired, with a row of melanophores
flanking both sides of the anal finfold (Foster, pers. comm.).
The early life history stages of phallostethid fishes follow closely
the atheriniform pattern. In both Gulaphallus mirahilts (Villa-
dolid and Manacop, 1934) and G. /a/a/f'r (Manacop, 1936) the
preanal length is short and a median series of melanophores
develops middorsally. The exact disposition of the dorsal me-
lanophores has not been described nor can it be assessed from
published illustrations.
Relationships
Two ontogenetic character states suggest that the atheriniform
fishes are a monophyletic group compwising an order, the Ath-
eriniformes, of equal standing with the Beloniformes and Cy-
prinodontiformes. First, the preanal length of all known ath-
eriniform flexion larvae, except Odontesthes dehueni, is short;
being approximately one-third of body length. Preanal length is
variable in the other two atherinomorph orders but the preanal
lengths of few, if any, beloniform or cyprinodontiform species
are this short between hatching and early flexion. The Perco-
morpha is thought to be the sister group of the Atherinomorpha
(Rosen and Parenti, 1981). In almost all primitive percomorphs,
preanal length exceeds that of the Atheriniformes through flex-
ion and approaches as much as 50-70% of body length (Ahl-
strom and Moser, 1976). The same can be said of the paracan-
thopterygian, myctophiform and aulopiform fishes (sensu Rosen;
1973, 1982). Preanal length is reduced in gadid fishes (Dunn,
this volume), but the short gut typical of the cods is always
looped and therefore is considered here to be nonhomologous
with the condition seen in the atheriniforms. Outgroup com-
parison thus suggests that the reduced larval preanal length can
Fig. 1 86. (A) Atheriniform eggs. Mature egg, Atherinops affinis. San-
ta Catalina Island, California. LACM field no. IP-77-43; (B) Ovarian
egg. Euryslole enarcha. LACM 31784-5; and (C) Atherinopsis califor-
niensis. egg. LACM 43446-1.
Fig. 187. Ovarian egg, Eurystole eriarcha. LACM 31784-5. (upper) lOOx; (lower) l.OOOx.
WHITE ET AL.: ATHERINIFORMES
359
Fig. 188. (A) Atheriniform larvae. Odomesthes debueni; 10.2 mm SL, from Fischer, (1963); (B) ho hawaiwnsis: 6.2 mm SL, from Miller et
al., (1979); (C) Athermomorus insularum; 5.4 mm SL, from Miller et al., (1979); and (D) Bedotia geayi; 5.3 mm SL, LACM uncatalogued.
be used as a synapomorphous character state to define the Ath-
eriniformes.
The second ontogenetic character stale suggesting that the
atheriniform fishes comprise a monophyletic group relates to
larval pigmentation and may contribute to their cladistic di-
agnosis. In all atheriniform larvae a single row of melanophores
develops on the dorsal margin (Fig. 189A). This situation con-
trasts with the Beloniformes and Cyprinodontiformes, where no
consistent larval pigmentation pattern is evident (Hardy, 1978a).
What is known of larval halfbeaks suggests that when a dorsal
pigment series occurs it is always composed of at least a double
row of melanophores (Fig. I89B).
While it is typical for cyprinodontiform larvae to develop
dorsal, lateral and ventral pigment series (Foster, 1967), no
consistent pattern is evident. In Fundulus. the middorsal me-
lanophores are arranged in a paired series (Hardy, 1978a). In
Cyprinodon vanegatus obscure blotches of pigment occur on the
body (Hardy, 1978a). Melanophores are evenly distributed over
the larva of Liuania parva (Hardy, 1978a). The larva of Epi-
platys sexfasciatus has melanophores randomly distributed over
its dorsal surface (Scheel, 1968). In the Atherinomorpha, only
the adrianichthyoid fishes have larvae with dorsal melanophores
arranged in a single row (Kulkami, 1940; Job, 1940). This re-
semblance to the Athcriniformes is considered to be convergent
because, given the mtemal relationships of the Atherinomorpha
(Rosen and Parenti, 1981), it is more parsimonious to assume
that a single dorsal melanophore row evolved independently in
the Athcriniformes and Adrianichthyoidei because only two
evolutionary events are involved. However, if this pigment pat-
tern is viewed as a sympleisiomorphy, it is necessary to invoke
360
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 94. Meristics of Selected Atheriniform Species. Only total fin elements are reported because of confusion in the literature as to the
proper definition of spines and rays and the inconsistencies in published descriptions that resulted from this confusion.
Fin
rays
Gill
Dl
D2
A
Pect.
Vert.
rakers
Reference
Alherinidae
Alia net la mugi hides
5-7
8-11
9-12
13-15
34-37
14-17
Ivantsoff,
1978
Alherina boyen
6-9
11-14
12-16
15-17
42-46
17-24
Ivantsoff,
1978
Atherina hepselus
7-9
10-13
11-14
17-19
52-55
25-28
Ivantsoff,
1978
Alherina presbyter
7-9
11-15
15-19
14-17
45-52
19-24
Ivantsofl^,
1978
Alherinason esox
5-9
10-15
11-15
12-14
45-48
18-24
Ivantsoff",
1978
Atherinason hepseloides
5-10
10-14
12-15
13-16
43-48
15-20
Ivantsoff,
1978
Alherinomorus capricornensis
3-7
10-11
13-16
15-18
42-44
20-24
Ivantsoff^,
1978
A therinomorus endrachtensis
4-6
9-11
12-15
13-17
35-39
18-22
Ivantsoff",
1978
Alherinomorus ogilbyi
4-7
9-12
12-17
16-19
37-42
22-28
Ivantsoff",
1978
Alherinomorus pinquis
4-7
9-12
13-16
15-19
38-43
18-25
Ivantsoflf,
1978
Atherinops affinis
3-7
10-14
9-14
12-15
43-49
14-27
White, unpub.
Atherinopsis californiensis
4-9
10-15
20-29
14-17
46-53
18-44
White, unpub.
Athennosoma elongala
4-8
9-13
9-15
12-16
37-43
12-16
Ivantsoff",
1978
Atherinosoma microstoma
5-9
9-12
9-14
12-16
37-42
12-15
IvantsoflT,
1978
Atherlnosoma presbyteroides
6-8
10-14
11-16
12-14
41-48
15-20
Ivantsoff",
1978
Athenon elymus
3-5
9-14
14-16
12-15
38-42
10-14
Ivantsoff",
1978
Atherion macculocht
3-5
10-13
15-19
12-14
43-46
10-16
Ivantsoff",
1978
Colpichthys regis
5-8
10-13
20-24
13-16
44-48
16-19
White, un
pub.
Craterocephalus cuneiceps
4-6
6-9
7-9
12-15
31-34
10-13
Ivantsoff,
1978
Craterocephalus dalhousiensis
4-6
6-8
8-10
13-15
30-32
8-9
Ivantsoff",
1978
Craterocephalus eyresll
3-6
7-9
7-11
12-14
32-41
10-13
Ivantsoff,
1978
Craterocephalus honoriae
5-6
8-9
10-13
13-15
35-38
12-15
Ivantsoff",
1978
Craterocephalus lacustris
5-8
8-10
9-11
13-17
35-39
10-13
Ivantsoff,
1978
Craterocephalus marjoriae
5-7
7-9
7-10
12-16
29-34
10-12
Ivantsoff,
1978
Craterocephalus nouhuysi
6-8
9-11
10-12
13-17
37-38
7-9
Ivantsoff",
1978
Craterocephalus pauciradiatus
4-6
7-9
8-11
12-14
30-35
10-13
Ivantsoff".
1978
Craterocephalus randl
5-8
8-10
8-12
13-15
34-39
7-11
Ivantsoff",
1978
C stercusmuscarum fulvus
4-8
7-11
8-11
13-16
31-36
10-13
Ivantsoff,
1978
C stercumuscaruin stercumuscarum
5-8
6-9
8-10
12-15
35-38
9-12
Ivantsoff",
1978
Hypoatherina barnesi
5-7
9-12
13-15
13-16
41-42
15-18
Ivantsoff",
1978
Hypoatherlna ovalaua
4-7
9-11
10-13
16-18
38-40
22-25
Ivantsoff,
1978
Hypoatherina temminckii
5-7
9-11
12-15
17-20
38-44
21-25
Ivantsoff",
1978
Hypoatherina Iropicalis
5-8
9-12
12-15
16-19
40-47
18-22
Ivantsoff,
1978
Hypoatherina valenciennsi
4-7
9-11
13-14
15-17
39-42
20-25
Ivantsoff",
1978
Leuresthes tenuis
4-7
9-13
20-24
13-16
47-50
20-29
White, un
pub.
Stenatherina panalela
6-7
9-10
10-13
17-19
21-24
40-45
Ivantsoff,
1978
Bedotiidae
Bedotia geayi
4-5
10-13
15-18
12
Pellegrin,
1907, 1914
Bedotia longianalis
5
14
20
12
Pellegrin,
1914
Bedotia madagascariensis
5
12
19
Pellegrin.
1914
Regan, 1903a
Isonidae
Iso hawaiiensis
Melanotaeniidae
Cairnsichthys rhombosomoides
Chilatherina campsi
Chilatherina crasslspinosa
Chilatherina lorentzi
Chilatherina sentaniensis
Glossolepis incisus
Glossolepis multisquamata
Glossolepis pseudoincisus
Iriatherina werneri
Melanotaenia affinis
Melanotaenia ajamaruensis
Melanotaenia hoesemani
Melanotaenia jluviatilis
Melanotaenia goldiei
Melanotaenia japenensis
Melanotaenia lacustris
Melanotaenia nigrans
Melanotaenia ogilbyi
4-5
17
23-25
12-13
5-6
14
19-21
4-8
13-17
21-25
4-5
9-13
21-25
4-7
13-18
24-31
4-5
10-15
23-26
5-6
10-11
21-24
4-5
9-12
18-22
5-6
11-13
19-23
13-14
6-9
8
11-13
4-5
15-19
21-25
4-6
16-20
22-28
13-15
4-6
11-15
18-24
13-16
5-7
13-14
19-21
13-15
5-6
12-16
21-25
4-5
16-18
27-29
13
4-5
12-14
19-20
4-7
10-14
18-22
5-7
10-12
18-19
35-38
Miller etal.. 197
9
36-37
10-12
Allen, 1980
13-14
Munro, 1967
14
Munro, 1967
15
Munro, 1967
13-15
Munro, 1967
32
Munro, 1967
16-19
Munro, 1967
26-30
Allen and Cross,
1980
32-33
11-13
Allen. 1980
13-17
Munro. 1967
14-15
Allen and Cross.
1980
14-15
Allen and Cross,
Scott etal., 1980
1980
14-16
Munro, 1967
13-14
Allen and Cross,
1980
14
Munro, 1967
13-15
Munro, 1967
12
Munro, 1967
WHITE ET AL.: ATHERINIFORMES
361
Table 94. Continued.
Fin rays
Dl
D2
A
Peel.
Ven
5-6
19-20
25-27
14-15
4-6
11-15
18-21
5-7
10-13
17-18
5-7
11-12
20-22
13-16
5-7
10-14
18-23
4-6
11-13
19-22
16-20
4-7
19-22
24-27
5-8
10-12
17-21
32
4-5
6-7
10-12
10-11
4-5
6-7
9-10
12-13
3-5
11-15
19-23
35
1-2
5-6
14-16
11
2
7
16-18
9
2
7-8
15-19
9
1-2
7
16-17
9-10
6-7
14-16
13
1-2
5-7
13-18
9
1-2
5
14-15
9
1-2
5-6
11-14
10
1
5-6
15-16
10
1
6-8
14-15
11
2
6
16
12
2
5
14-16
9
1-2
5-6
8-10
15-17
26-28
9
6
14
10
35
5
14
9
36
2
5
14-16
2
5-7
10-11
8-10
19-23
12-13
Gill
rakers
Melanotaenia oktediensis
Melanotaenia praecox
Melanotaenia sexlmeata
Melanotaenia splendida australis
Melanotaenia s. rubrostnata
Melanotaenia trifasciata
Melanotaenia vanheurni
Popondelta furcatus
Pseudomugil gertrudae
Pseudomugil tenellus
Rhadinocentnts ornalus
Phallostethidae
Ceratostethus bicornis
Gulaphallus eximius
Gulaphallus mirabilis
Manacopus falcifer
Mirophallus bikolanus
Neostethus amaricola
Neostethus borneensis
Neostethus coronensis
Neostethus lankestn
Neostethus panayensis
Neostethus siamensis
Neostethus villadolidi
Neostethus zamboanga
Phallostethus duncken
Phenacostethus smithi
Phenacostethus posthon
Plectrostethus palawanensis
Solenophallus ctenophonis
Solenophallus thessa
Telmatherinidae
Tehnathenna celebensis
6-7
11-12
13-15
16
32
15
Allen and Cross, 1980
11-12
Munro, 1967
12
Munro, 1967
16
Taylor. 1964
14-16
Munro, 1967
Taylor, 1964
14
Munro, 1967
8-10
Allen, 1980
Taylor, 1964
9
Taylor, 1964
11-12
Allen, 1980
Herre, 1942
Herre, 1942
12-13
Herre, 1942
Herre, 1942
12
Herre, 1942
Herre, 1942
Herre, 1942
10-11
Herre, 1942
Herre, 1942
15
Herre, 1942
Herre, 1942
15
Herre, 1942
12
Herre, 1942
Regan, 1913c
Roberts, 1971
Roberts, 1971
Herre, 1942
13
Herre, 1942
16
Herre, 1942
Boulenger, 1897
Fig. 189. Dorsal pigment series in athennomorph larvae. (A) Atherinops affinis; 6.7 mm SL, LACM field no. 42841-3; and (B) Dermogenys
pusdlus: 7.7 mm SL, LACM 43448-1.
362
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
Table 95.
Derived Developmental Character States in the
Atherinomorpha.
Alhenni-
formes
Cypnno-
Beloni- donti-
formes formes
Egg large, demersal with chorionic fila-
ments and lipid globules coalescing at
vegetal pole
Separation of embryonic afferent and effer-
ent circulations by development of heart
in front of the head
Formation of spermatogonia near the tuni-
ca albuginea
Gut length less than 40% of flexion length
Single row of melanophores on dorsal sur-
face
Fin rays present at hatching
X
X
the development of a single dorsal melanophore row in the
common ancestor of the Atherinomorpha and separate loss
events in the Exocoetoidei and Cyprinodontiformes.
' After this paper went to press, a report on the relationships of the
phallostethid fishes appeared in which Parenti (1984) questions our
conclusions on atherinifomi monophyly. The evidence she presents
suggests another phylogenetic interpretation but at the present time
neither hypothesis can definitely be rejected.
Aspects of the variable reproductive behavior of the ather-
iniform fishes might be useful in determining relationships with-
in the order. The habit shared by Leuresthes tenuis and L. sar-
dina of spawning on the beach in synchrony with the lunar cycle
(Thompson and Thompson, 1919; Clark, 1925; Walker, 1952)
is a synapomorphy identifying the two grunion species as each
other's closest relative. Another lunar spawner, Menidia men-
idia deposits its eggs in detrital mats and on the stems and
exposed roots of the cordgrass plant Spartina alternijlora (Moore,
1980; Middaugh et al., 1981). Telmatherina ladigesi deposit
their eggs over a period of several days, attaching them singly
and in a widely spaced pattern to aquatic vegetation. In the
Phallostethidae, fertilization is internal and the eggs are attached
to a substratum by their adhesive filaments. There is much
variation in the reproductive behavior of atheriniform fishes
and investigation of their breeding habits might secure infor-
mation bearing on their systematic relationships.
Table 95 summarizes the derived ontogenetic characters that
bear on atherinomorph relationships. There is still much that
is unknown about the early life history of the atheriniform fishes
and it is reasonable to hope that future investigation, particularly
of their reproductive habits and egg morphology, will contribute
to the elucidation of their evolutionary relationships.'
Section of Fishes, Natural History Museum, Los Angeles,
California 90007.
Cyprinodontiformes: Development
K. W. Able
THE approximately 800 members of the Cyprinodontiformes
(killifishes) are small to medium-sized fishes (8-300 mm
SL) that live in shallow fresh and brackish water. They are nearly
worldwide in their distribution in temperate and tropical areas
(Parenti, 1981). Cyprinodontiformes is considered to be mono-
phyletic based on several adult osteological characters and the
long embryonic development time (Parenti, 1 98 1 ). I here follow
the most recent and extensive revision of the group by Parenti
( 1 98 1 ) in which she rearranges them into two suborders: Aploch-
eiloidei with 2 families (Aplocheilidae and Rivulidae) and Cy-
prinodontoidei with 7 families (Profundulidae, Fundulidae,
Valenciidae, Anablepidae, Poeciliidae, Goodeidae, Cyprino-
dontidae). See Nelson (1976) and Parenti (1981, Table 3) for
prior classification schemes. Comments on portions of Parenti's
reclassification can be found in Klee (1982) and Foster (1982).
Reproduction and development within the group is excep-
tionally varied, with oviparity, ovoviviparity, viviparity (in-
cluding functional states of each) and functional hermaphro-
ditism represented. In addition, viviparity may have evolved
independently at least four times within the order (Parenti, 1981).
Among the viviparous forms occur a vast array of schedules
and morphological modifications for internal development such
as the trophotaeniae of the goodeids and the intra- and extra-
follicular gestation and superfetation in some poeciliids. De-
velopment reportedly is long, from four days to more than one
year (Scheel, 1962) in some of the "annual" killifishes. The
rivulid, Rivulus mannoratus, is unique among fishes, and ver-
tebrates in general, in that it is a functional hermaphrodite with
internal fertilization (Harrington, 1961). Published early life his-
tory descriptions are listed in Table 96.
Eggs
The eggs of some Cyprinodontiformes are among the smallest
known for fishes. Scrimshaw (1946) recorded fertilized eggs of
the poeciliid Hcterandria fonnosa. in which development is
internal, to average 0.30 mm and Roberts (1970) recorded "ripe"
eggs of another poeciliid, Fluviphylax, as 0.1 mm (not substan-
tiated). The eggs of other cyprinodontiforms are larger (Table
97) with the largest that of Fiindiilus majalis at 2.0-3.0 mm.
Egg size varies within some species (i.e., the cyprinodontid
Aphanius anatoliae. Grimm, 1979a, b) and judging from the
data in Table 97, may vary in other species as well. Other authors
have noted differences in the egg size of cyprinodontids of the
genus Cyprinodon and considered them to be environmental
(Soltz and Hirschfield, 1 98 1) or genetic (Garrett, 1 982). Fecun-
dity is correlated with egg size in the aplocheilid Nothohranchius
(Bailey, 1972) and with female size in poeciliids (see Thibault
and Schultz, 1978). Fecundity also varies between females and
populations in the oviparous goodeid, Crentchthys baileyi (Es-
pinosa, 1968). Superfetation occurs in several poeciliid genera
ABLE: CYPRINODONTIFORMES
363
(Turner, 1937; Scrimshaw, 1945; Turner 1940a; Thibault and
Schultz, 1978).
The eggs of all cyprinodontiforms contain conspicuous oil
droplets (Foster. 1967) (Table 97) including the viviparous poe-
ciliids such as Gambusia affinis (Kuntz, 1914a). Within the
Fundulidae the size and number of oil droplets is extreme; Lu-
cania parva has 8-12 large droplets (Fig. 190C) and Fundulus
n. sp. from Bermuda has up to approximately 350 with a mean
of 181 droplets per egg (Fig. 190F, and Able et al., in prep.).
Subspecific variation in the fundulid F. heteroclitus is pro-
nounced and population means range from 10 to 180 droplets
(Morin and Able, 1983).
These droplets probably provide nutrition late in embryonic
development (Smith, 1957; Lentz and Trinkhaus, 1967; Blaxter,
1969a; Temer, 1979). The chemical composition of lipids in
the oil droplets has been determined by Bailey (1973). The oil
droplets are clumped together at ovulation but disperse after
fertilization. Individual oil droplets are retained in the yolk sac
after hatching in several Fundulus species, L. parva (see Hardy,
1978a), R. marmoratus (McMillan, 1979) and postflexion G.
affinis (Ryder. 1885). The eggs of all known oviparous and ovo-
viviparous cyprinodontiforms have a small perivitelline space
and are spherical (except in Nothobranchius in which the egg is
oval, Scheel, 1968).
The chorion is variable in thickness and surface structure
(Table 97). In most of the oviparous and ovoviviparous forms
the chorion is multilayered and thick, whereas in many vivipa-
rous forms it is considerably reduced (see Flegler, 1977). De-
tailed studies of the chorion microstructure are available for the
fundulid F. heteroclitus (Kuchnow and Scott, 1977) and the
rivulid Cynolebias bellottii (Sterba and Muller, 1962; Muller
and Sterba, 1963). The chorion of all oviparous and ovovivipa-
rous forms have adornments of some type on the surface. In-
stances where they have been reported as lacking (F. heteroclitus,
F. parvipinnis. Foster, 1967; F. majalis. Hardy, 1978a) are in-
correct. Often the chorion is covered with filaments either uni-
formly arranged or clustered together to form tufts (Fig. 190,
191; Table 97). The filaments can vary in diameter and density
between species (Fig. 190 and 191) and subspecies (Dumont
and Brummetl, 1980; Morin and Able, 1983). Differences in
these structures in F. heteroclitus appear to be correlated with
spawning site preference (Able, 1984). Fundulus majalis has
microfilaments on the large filaments and on the chorion surface
(Fig. 191 A, B). Some species have other structures ("punctae"
of Foster, 1967) which appear as small spherical knobs on the
surface of the chorion, occasionally with filaments originating
from them (Fig. 190D, E; 19 ID, E). In other species the surface
of the chorion may be sculptured (Table 97). Fundulus luciae
has numerous circular pits in the chorion surface (Fig. 190D,
E). The distribution of chorionic modifications (filaments, mi-
crofilaments, pits, knobs) within the Cyprinodontiformes is in-
completely known and thus it is difficult to assess their phylo-
genetic significance. Several species of fundulids studied possess
punctae or knobs (Table 97; Figs. 190, 191) while these are
lacking in the cyprinodontids (see Fig. 1 90A, B) thus supporting
the separation of these groups by Parent! (1981).
The presence of chorionic filaments in the Cyprinodonti-
formes is a synapomorphy shared with the Atheriniformes and
Beloniformes as discussed in this volume and constitutes one
of the synapomorphies serving to unite the Atherinomorpha.
Further studies of egg morphology in the oviparous forms will
Table 96.
Published Descriptions of Cyprinodontiform Early Life
History Stages Listed by Family and Genus.
Family and genus
Sources
Suborder Aplocheiloidei
Aplocheilidae
Fundulopanchax
Peters, 1963
Nothobranchius
Peters, 1963
Zahradka and Frank, 1976
Rivulidae
Rivulus
McMillan, 1979
Suborder Cyprinodontoidei
Profundulidae
None
Fundulidae
Planet erus
Koster, 1948
Fundulus
Ryder, 1885
Kuntz, 1914a. 1916
Fish, 1932
Armstrong and Child, 1965
Foster, 1967. 1974 '
Byrne, 1978
Hardy, 1978a
Jones and Tabery, 1980
Lucania
Kuntz, 1916
Foster, 1967, 1974
Hardy. 1978a
Leptolucania
Foster. 1967
Adinia
Foster, 1967
Koenig and Livingston, 1976
Valenciidae
None
Goodeidae
Crenichthys
Kopec, 1949
Cyprinodontidae
Cubanichthys
Troemner, 1932, 1941
Cyprinodon
Kuntz. 1916
Foster, 1974
Hardy, 1978a
Mettee and Beckham, 1 978
Jordanella
Foster, 1967
Anablepidae
Anableps
Turner, 1940c
Jenynsia
Turner, 1940d
Poeciliidae
Gambusia
Ryder, 1885
Kuntz, 1914a
Xiphophorus
Tavolga, 1949
probably provide useful insights into the phylogeny of this abun-
dant and diverse group.
Embryonic development
Embryonic development within the Cyprinodontiformes is
almost as variable in duration and number of physiological and
morphological modifications as in all other fishes combined.
The incubation time may be as short as 4-8 days in C. variegatus
and Jordanella Jloridae (see Foster, 1967; Hardy, 1978a) to
possibly longer than a year in some of the "annual" species.
Parent! (1981) considers this annual habit to have developed
more than once within the Cyprinodontiformes. This is sup-
364
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 97. Summary of Ego Characteristics of Cyprinodontiform Fishes.
Species
Egg diameter
(mm)
Oil globules
Chorion surface
Source(s)
Suborder Aplocheiloidei
Family Rivulidae
Cynolebias bellotlii
Cynolebias ladigesi
Cynolebias whilei
Cynolebias melanotaenia
Procatopus nototaenia
ca. 2.0
7
ca. 1.0
7
1.0
present
7
present, large
7
9
anchorlike filaments
prickles
palm like stems
filaments
Breder and Rosen, 1966
Peters. 1963
Breder and Rosen, 1966
Scheel, 1962
Scheel, 1961
Family Aplocheilidae
Fundulopanchax amieti
Fundulopanchax arnoldi
Aplocheilus blocki
Aplocheilus Imeatus
Epiplatys chaperi
Epiplalys senegalensis
Nothobranchius guentheri
Nothobranchius korthausae
7
7
13
2.0
•>
1.0
9
0.9-1.1
?
?
7
9
present, large
7
present, large
present, large
filaments
hexagonal pattern
3 long filaments
7
filaments in tuft
7
filaments
filaments
Carr, 1982
Peters, 1963
Jones. 1937
Job, 1940
Peters, 1963
Breder and Rosen, 1966
Peters, 1963
Zahradka and Frank, 1976
Suborder Cyprinodontoidei
Family Fundulidae
Planclerus kansae
Fundulus chrysotus
F. cingulatus
F. confluentus
F. diaphanus
F. heteroclitus
F. luciae
F. majalis
F. notatus
F. nottii
F. olivaceus
F. parvipinnis
F. similis
Lucania goodei
Lucania parva
Leptolucania ommata
Adinia xenica
Family Poecilidae
Gambusia ajjinis
Poecilia reticulata
Poeciliopis lucida
P. monacha
P. prolifica
P. lurneri
Tomeurus gracilis
Family Goodeidae
Crenichthys baileyi
Family Valenciidae
Valencia hispanica
2.3-2.4
7
filaments, punctae
Foster, 1967
2.0-2.1
7
filaments, punctae
Foster, 1967
1.5-1.6
7
filaments, sculptunng
Foster, 1967
1.6-1.8
10-15, medium
filaments, punctae
Harrington, 1959; Foster,
1967; Hardy, 1978
1.7-2.4
10-15. medium and
filaments, punctae
Foster, 1 967; Wang and
40-80, small
Kemehan, 1979
1.5-2.5
10-180
filaments, variable
Hardy, 1978a; Morin and
Able. 1983
1.7-2.2
5-58
filaments, punctae
Hardy, 1978a
2.0-3.0
50
absent ? and present
Ryder, 1885; Nichols and
Breder, 1927; Hardy, 1978a
Wang and Kemehan, 1979
ca. 1.8
many, medium also
filaments in tuft
Foster, 1967; Jude, 1982a
single
and punctae
2.2-2.3
numerous
filaments
Foster, 1967
ca. 1.8
many, medium
filaments in tuft
and punctae
Foster, 1967
2.8
present
filaments
Ritterand Bailey, 1908;
Hubbs, 1965
2.6-2.9
7
?
Foster, 1967; Martin and
Finucane, 1969
1.3
10-12, medium
filaments in tuft
Foster, 1967
1.0-1.3
10-12(0.3-0.4 mm)
filaments in tuft
Kuntz, 1916; Foster, 1967;
Hardy, 1978a
1.0-1.1
7
filaments in tuft,
sculpturing
Foster, 1967
1.5-2.2
many, small
filaments
Foster, 1967; Koenig and
1.6-2.1
present
1.7
9
1.4
9
2.0
7
1.0
7
1.0
7
9
9
absent
7
7
7
7
7
many, long filaments
1.9-2.0
2.5-2.6
Livingston, 1976; Hastings
and Yerger, 1971
Hardy, 1978a
Thibaultand Schultz, 1978
Thibault and Schultz, 1978
Thibaultand Schultz, 1978
Thibaultand Schultz, 1978
Thibault and Schultz, 1978
Breder and Rosen, 1 966
Kopec, 1949; Bnll, 1981
Rachow, 1924; Villwock,
1960
ABLE: CYPRINODONTIFORMES
365
Table 97. Continui
;d.
Egg diameter
Species
(mm)
Oil globules
Chorion surface
Source(s)
Family Cyprinodontidae
Cubanichlhvs cuhensis
ca. 1.0
?
present
Mayer, 1933
Cubanichlhvs pengellevi
1.2-1.4
few
filaments in tuft
Foster, 1969
Aphamus fasciatus
7
7
filaments
Breder and Rosen, 1 966
Aphanius memo
1.4-1.5
ca. 6, medium
filaments
Mazza, 1902
Cyprinodon macularius
ca. 2.0
7
7
Constantz, 1981
Cvprinodon ncvadensis
1.3-1.4
7
7
Constantz, 1981
Cyprinodon variegatus
1.1-1.7
one large
many minute
filaments
Foster, 1967; Fanara,
1964; Wickler, 1959
Floridichthys carpio
1.4
7
1-2 filaments, long
Kaill, 1967
Jordanella Jloridae
1.3-1.4
7
filaments
Henzelmenn, 1930;
Foster. 1967; Kaill, 1967
Jordanella pulchra
1.0
7
7
Cassel, 1981
'^^^^^^^^^^^^K^
■-L-
^ "^WHR
^^^if^
■''^■^-'^^^*^
Fig. 190. Scanning electron micrographs of the chonon surface of the cypnnodonlid Cyprinodon aharezi (A, B) and the fundulid Fundulus
luciac (D, E). Oil droplets are shown for the fundulids Lucania parva (C) and Fundulus n. sp. from Bermuda (F).
366
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
Fig. 191. Scanning electron micrographs of the chorion surface of the fundulids F. majalis (A, B), Fundulus n. sp. from Bermuda (C), the
rivulid Rinilus mannoratus (D, E) and the profundulid Profundulus punctalus (F).
ported by an apparent difference in the manner in which hatch-
ing is delayed. Fundulus confluentus can hatch after three months
of "latency" or postponement of hatching (Harrington, 1959;
Harrington and Haeger, 1958) while the embryos continue to
grow and utilize yolk reserves. Delayed hatching is probably
typical for many North American fundulids, as seen in F. het-
eroclitus (Taylor et al., 1977) and Adinia xenica (Koenig and
Livingston, 1976). The incubation period is known to be influ-
enced by temperature (see Gabriel, 1944) and dissolved oxygen
(DiMichele and Taylor, 1980). During diapause, which occurs
in the annual killifishes (Wourms, 1972a, b, c) hatching may be
delayed for up to six months in nature and possibly longer than
a year under extreme conditions. During this time growth does
not occur, cardiac activity ceases and the yolk is not depleted.
The length of the incubation period may be controlled by tem-
perature, photoperiod, desiccation and oxygen tension cues (see
Matias, 1982).
The embryonic development of several aplocheilids (Aploch-
eilus), a rivulid {Rivulus) and two fundulids (Adinia and F.
heteroclitus) has been described in detail (Table 96). Some
authors have placed special systematic significance on the pat-
tern of vitelline circulation of the embryo in cyprinodonti forms
(Foster, 1967; Hubbs and Bumside, 1972) and other atherino-
morphs (White et al., this volume). The viviparous poeciliids,
anablepids, jenynsiids (placed in the Anablepidae by Parenti,
1981) and goodeids have a variety of modifications for receiving
nourishment during development (reviewed by Wourms, 1981).
The phylogenetic significance of independent development of
viviparity in several cyprinodontiform lineages is discussed in
detail by Parenti (1981).
Larvae
The larvae of oviparous cyprinodontiforms are incompletely
known (Table 96) despite the fact that many of them are avidly
bred by aquarium hobbyists. All of those known lack the preanal
finfold characteristic of the beloniforms (except exocoetids)
(Collette et al., this volume) and have a longer preanal length
than the atheriniforms (White et al.. this volume). In all cy-
ABLE: CYPRINODONTIFORMES
367
Mur
^(S;^
Fig. 192. Larvae of (A) the aplocheilid Nothobranchius eggersi, 3.1 mm SL; (B) the nvulid Rivulus marmoratus, 4.6 mm SL; and (C-E) the
fundulid Funduliis n. sp. from Bermuda. 6.0 mm SL.
368
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
prinodontiforms that have been studied the caudal fin rays form
first (Fig. 192) and often this occurs before hatching (Foster,
1967).
Within the oviparous cyprinodontoid killifishes the presence
and location of melanophores as well as the relative location of
the dorsal finfold may be useful systematic characters (Foster,
1968). In most fundulid larvae the dorsal finfold originates pos-
terior to the origin of the anal finfold (Foster, 1967; Fig. 192)
with the possible exception of Lucania parva (see Hardy, 1 978a).
In the cyprinodontids studied however, the dorsal finfold orig-
inates anterior to the anal finfold (Foster. 1967). The larvae of
most fundulids, the aplocheilid, Nothobranchius eggersi and the
rivulid, R. marmoratus also possess three rows or stripes of
melanophores (middorsal, midlateral and midventral) on the
body (Fig. 192). This characteristic is shared by the beloniform
Oryzias latipes and some atheriniforms (Martin and Drewry,
1978) and suggests that this character may be symplesiomorphic
within the Atherinomorpha. In cyprinodontids these rows of
melanophores are lacking and the existing melanophores are
scattered evenly over the body or appear as saddle-shaped groups
of melanaphores on the dorsolateral surface of the body (see
Foster, 1967; Hardy, 1978a).
Summary
The early life history of cyprinodontiforms appears to offer
many possibilities for elaborating on their phylogeny. Several
authors (Rosen and Parenti, 1981; Collette et al., this volume;
White et al., this volume) have pointed out the usefulness of
early life history characters in defining the monophyletic nature
of the Atherinomorpha and the orders within. Although studies
of the early life history of the Cyprinodontiformes are not as
far along, they may offer more potential for several reasons.
First, Foster (1967, 1968) has already pointed out the value of
early life history characters in resolving the phylogeny of the
group. Second, based on this review, both egg morphology and
larval characters vary within the group and thus seem to offer
real promise for assessing relationships. Third, many killifishes
are easily maintained and will reproduce in aquaria so that study
material should be easily obtainable, especially given their pop-
ularity in the aquarium trade.
Biological Sciences and Center for Coastal and Envi-
ronmental Studies, Doolittle Hall, Rutgers
University, New Brunswick, New Jersey 08903.
Lampriformes: Development and Relationships
J. E. Olney
THE order Lampriformes (=Lampridiformes, see Robins et
al., 1980) is composed of approximately 21 species (Table
98) of pelagic, marine fishes with worldwide distribution ex-
cluding polar seas. Highly evolved and extremely divergent in
form and lifestyle, these species occupy meso- and epipelagic
habitats and have attained a remarkable degree of specialization,
of which the most notable examples are: the pectoral muscu-
lature of Laniphs (Rosenblatt and Johnson, 1976); the unique
feeding mechanism of S7r/e/'/7C'n«(Pietsch, 1978a); the ribbon-
like body form, specialized integument and rotating eye of the
trachipterids (Walters, 1963; Haedrick, 1974; Oelschlager,
1976a); the "horn" of Eiimecichthys (Fitch, 1966; Oelschlager,
1979); and the cephalopod-like ink gland of the lophotids and
Radiicephalus (Walters and Fitch, I960; Fitch and Lavenberg,
1968; Harrison and Palmer, 1968; Saldanha and Pereira, 1977;
and others). By far the most impressive species of the group is
the oarfish, Regalecus glesne. which attains lengths of over 8m,
possesses a crimson dorsal fin and cockscomb-like anteriormost
dorsal rays and is the probable basis for many historical sightings
of sea monsters (Fitch and Lavenberg, 1968).
Regan (1907, 1924) first suggested a relationship between
Lophotus. Eumecichthys, Lampris, V'elifer and Stylephorus. all
on the basis of the common possession of peculiar characteristics
of the protractile mouth and assigned these genera to a new
order, the Allotriognathi (from the Greek, meaning "strange
jaw"). Presently, the order consists of 12 genera (Velifer. Me-
tavelifer, Lampris. Zu, Desmodema, Trachipterus, Radiicepha-
lus, Lophotus. Eumecichthys. Stylephorus. Regalecus and
Agrostichthys) comprising seven families (Table 98).
Two conflicting proposals exist for the allocation of these
fishes and nomenclatural inconsistencies persist. Oelschlager
(1976a,b, 1978a, b, 1979; also see Palmer, 1973) retains Regan's
( 1 907) ordinal designation and defines two suborders of the
Allotriognathi: the Bathysomi, deep-bodied fishes with sym-
metrical caudal fins, well developed skeletons and musculature
(represented by Lampris and the veliferids); and the Taenio-
somi, elongate fishes with asymmetrical caudal fins, weak skel-
etons and musculature (represented by Trachipterus. Regalecus
and remaining genera). In contrast. Greenwood et al. (1966)
recognize four suborders of the Lampriformes: Lampridoidei,
Veliferoidei, Trachipteroidei and Stylephoroidei. At lower taxo-
nomic levels, Heemstra (in press) considers Lophotus to be
monotypic while Briggs(l952)and Oelschlager (1 979) recognize
two species (Table 98). In addition, a number of nominal species
exist within the genera Regalecus. Trachipterus and Lampris.
Recently, Heemstra (in press) and Heemstra and Kannemeyer
(in press) have treated South African Lampriformes, describing
a new Zu species and providing synonymies of several trachip-
terids. In general, the systematic status of lampriform fishes is
in question and the nomenclature lacks stability owing, in part,
to the rarity of examples in systematic collections and the fragile
nature of these fishes.
OLNEY: LAMPRIFORMES
369
Table 98. Recorded Meristics of Adult Lampriform Fishes. Total element counts are reported without reference to ray or spine designa-
tion of onginal source. Abbreviations used are: ABS— absent; PRE— precaudal vertebrae; CAUD— caudal vertebrae; TV — total vertebrae;
DORS— dorsal fin rays; ANAL— anal fin rays; PECT— pectoral fin rays; PELV— pelvic fin rays; CAUD— caudal fin rays.
pre
CAUD
TV
DORS
ANAL
PECT
PELV
CAUD
References
Veliferidae
Velifer hypselopterus
Metavelifcr nnilliradialus
16
16
17-18
17-18
33-34
33-34
34-36
41-45
25-26
33-36
15-16
15-16
7-8
9
36
36
Heemstra (in press); Regan
(1970); Walters (1960)
Heemstra (in press); Ste-
phenson (1977); Walters
(1960)
Lampridae
Lampris guttatus
L. immaculatus
Trachipteridae
7.U cristatus
T. jacksonensis
Radiicephalidae
Radiicephalus elongalus
Lophotidae
Lophotus lacepedei
L. capcUci
Eunwcichthys fiski
Stylephoridae
Stylephorus chordatus
Regalccidae
Regalecus glesne
Agrostichthys parkeri
21
25
46
43
48-52 33-42 21-24 13-17
52-56 35-38 23-24 12-14
22-24 39
Z. elongatus
29-31
—
84-87
Desmodema polysliclum
18-20
-
71-74
D. lorum
21-25
-
106-111
Trachiplcrus fukuzaki
25-28
-
69-72
T. allivetis
35-40
90-94
T. Irachiplerus
35-39
—
84-96
T. arcticus
99-102
T. ishikawae
—
—
73
62-69 120-150 ABS 10-12 3-7
142-147 ABS 11-12 7
120-124 ABS 12-14 ABS
197 ABS 12-14 ABS
153-174 ABS 11-13 5
165-184 ABS 10-11
145-185 ABS 9-11
133-168 ABS
9-11
13
31-34
81-83
166-173 ABS 13-14
36-39 77-79 114-121 152-160 6-7
9-10
170
6-7
5
5-6
_ ,_ 124-153 206-263 12-20 14-17 5
- - - 238 18 14-16 6
56 101 151-200 310-392 5-9 13-15 2-3
- - 50 115-122 16-17 10-11 -
_ _ 143-151 260-412 ABS 12-13 1
400 ABS 8-11 1
30-32 Herald (1939); McKenzie
and Tibbo (1963),
Regan (1907a)
— Parin and Kukuyev (1983)
8-12 + 1-5 Fitch (1964); Heemstra
and Kannemeyer (in
press); Palmer (1961 );
Walters and Fitch (1960,
1964)
12 + 5 Heemstra and Kannemeyer
(in press)
7-10 Rosenblatt and Butler
(1977)
4-7 Rosenblatt and Butler
(1977)
7-9 + 6-7 Fitch (1964,. 1967); Walters
and Fitch (1964)
7-8 + 6 Fitch (1964)
8+5 Heemstra and Kannemeyer
(in press); Palmer (1961)
8 + 5-6 Palmer (1961)
— Heemstra and Kannemeyer
(in press); Nishimura
(1964); Smith (1956a)
— Heemstra and Kannemeyer
(in press)
4 + 7 Harrison and Palmer
(1968); Heemstra and
Kannemeyer (in press);
Karrer(1976)
16-17 Heemstra (in press); Sal-
danha and Pereira
(1977)
17 Briggs (1952); Coin and
Erdman(1951)
12-13 Abe (1954); Fitch (1966);
Heemstra (in press);
Walters and Fitch (1960)
5-6 + 2 Pietsch (1978a); Regan
(1924)
3-4 Heemstra (in press); Tru-
nov (1982); Walters and
Fitch (1960)
2 Heemstra (in press); Tru-
nov (1982); Walters and
Fitch (1960)
F.g. 193. Eggs and larva (A-C) of Trach.pterus sp. (larva, 7.6 mm NL) after Mito (1961b). (D) Larva of Lophotus sp. (12.1 mm NL) after
Sanzo(1940).
OLNEY: LAMPRIFORMES
371
Development
Walters and Fitch (1960), Breder and Rosen (1966), Palmer
(1973), Nielsen (1973) and Moser (1981) have summarized the
state of knowledge of the early life history of lampriform fishes.
Eggs and larvae of the Veliferidae, Radiicephalidae and Style-
phoridae are unknown, although Karrer (1976) has mentioned
ripe ovarian eggs of Radiicephalus clongatus. Harrison and
Palmer (1968) presented meristic and morphometric data on a
154 mm SL R. elongatus termed a postlarva and Regan (1924)
figured a 26 mm SL larval Stylephorus chordatus. Little data on
young stages of the Lampridae are available. Ehrenbaum ( 1 905-
1 909) and Gudger ( 1 930) presumed pelagic eggs based on ovar-
ian examination; Gudger (1930), D' Ancona (1933b) and Oelsch-
lager (1976b) figured juvenile stages of Lampris giutatus; and
Parin and Kukuyev (1983) illustrated a young specimen of L.
irnmaculatus. Within the Lophotidae, larvae of Eumecichthys
are unknown while Fitch (1966) reported on ovarian eggs in E.
fiskt and Parin and Pokhilpkaya (1968) figured juvenile stages.
Sanzo ( 1 939b, 1 940) and Sparta (1954) have described eggs and
early larvae of a species oi Lophotus considered by Oelschlager
(1979) to be L. lacepedei. Eggs and larvae of trachipterid and
regalecid fishes have received considerable attention although
early life history stages of Agrostichthys and Desmodema are
unknown. Eggs and early larvae of Zu cristatus were described
by Sanzo (1918), Sparta (1933) and Olney and Naplin (1980).
Eggs and larvae of Trachipterus, probably representing several
species, were described by Emery (1879), Lo Bianco (1908a),
Jacino (1909), Ehrenbaum (1905-1909), Sparta (1933), Mito
(1961b) and Sardou ( 1 966). Eggs and larval stages of Regalecus
were figured and described by Sanzo (1925), Sparta (1933) and
Robertson (1975a). In summary, published information on the
development of eggs and larvae of four of the 1 2 lampriform
genera is available. In the following discussion, these published
data as well as additional material are utilized to summarize
the important characteristics of eggs, larvae and young of lam-
priform fishes and provide illustrations of larvae of four addi-
tional genera.
Egg and embryonic morphology. — Data on morphology and de-
velopment of lampriform eggs are incomplete (Table 99) but
indicate that eggs are large (1.7-4.0 mm egg diameter, range
excludes measurements of ovarian eggs,-see Table 99), spherical,
pelagic, often brightly colored (generally in amber, pink or red
hues) and possess thick, resilient chorions. Up to three weeks
may be required in incubation ( 1 8-20 days for R. glesne, Sparta,
1933). As a result, eggs are distinctive and easily recognized in
plankton collections (Fig. 1 93B, C) especially in advanced stages
of development (Orton, 1955a; Olney and Naplin, 1980).
Sanzo (1940) reported both homogeneous (Lophotus) and
segmented yolks (Zu and Regalecus) but recent observations
indicate homogeneous yolks in all known forms (Robertson,
1975a; Olney, unpublished data). Egg diameters, presence or
absence of oil droplets, chorionic ornamentation and micro-
structure may delimit some species (Table 99). Scanning elec-
tron micrographs of cross-sections of the chorions of Zu cris-
talus and an unidentified trachipterid species (Fig. 194) indicate
variability in chorion thickness and layering which may be of
systematic value. In general, however, confirmed identification
of lampriform eggs requires late stages with advanced embryos
(Olney and Naplm, 1980).
™* JM
[Hi
m
■1
K
leKv
g_
^mm^^
^■i
12138
VMS
X26e0
10U
565
Fig. 1 94. Scanning electron micrographs of chorionic microstructure
in lampriform eggs. (A) trachipterid, chorion thickness 1 1.04 ^m. (B)
Zu cristatus, chorion thickness 13.3 ^m.
Lampriform embryos exhibit precocious development. In
Trachipterus. Zu, Lophotus and Regalecus. anterior dorsal rays,
pelvic rays, distinctive pigment and total myomeres are appar-
ent and distinguish these forms (Sparta, 1933; Mito, 1961b;
Sanzo, 1940; Olney and Naplin, 1980). Some disparity exists,
however, in descriptions of late embryos. Sparta (1933) depicts
late embryonic R. glesne with anterior elements reduced but the
fourth elongate while Robertson (1975a) figures R. glesne em-
bryos off New Zealand with an elongate first element followed
by three reduced rays.
Larval morphology. — Al hatching, larvae of lampriform fishes
possess a number of distinctive characteristics including; well
developed, protrusible jaws; diflierentiated guts with an open
lumen and little or no yolk material; elongate anterior dorsal
elements which insert between the posterior eye margin and the
shoulder and are usually ornamented with broad, spatulate and
372
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 99. Some Characteristics of Eggs of Lampriform Fishes.
Egg
diameler
Descriplive
Species
(mm)
features
Reference
L. gultatus
—
Ovarian; thick
chorion with
amber tint
This Study
Z. cristatus
2.17-2.27
Thick chorion
Sanzo, 1918;
with amber tint;
Olney and
oil globule ab-
Naplin, 1980
sent
Trachipterus sp.
1.78
Yolk with scat-
tered melano-
phores
Mito. 1961b
T. trachipterus
2.9-3.2
Thick chorion; oil
globule absent;
yolk scattered
melanophores
Sparta, 1933
R. glesne
2.4-2.5
Numerous, scat-
tered oil glob-
ules; yolk with
scattered mela-
nophores
Sanzo, 1925
R. glesne
3.25-4.05
Chorion with pink
Robertson,
tint
1975a
Lophotiis sp.
2.48-2.64
Oil globules ab-
sent; chorion or-
namented with
small spines
Sanzo, 1940
E.fiski
1.5-2.0
Ovarian; transpar-
ent
Fitch, 1966
Table 100. PatternsofPterygiophoreInterdigitationin Anteri-
or Interneural Spaces of Young Lampriform Fishes. Within inter-
neural spaces, P indicates a predorsal element and numerals indicate
numbers of pterygiophores.
highly pigmented serial swellings; stout, well developed pelvic
eleinents often with fleshy sheaths and highly pigmented ter-
minal swellings; and snout to vent lengths approximately 40-
60% NL (Figs. 193, 195). Larvae of Metavelifer and Lampris
are slender at small sizes but Lampris larvae (and presumably
veliferids) rapidly increase in body depth (Fig. 196). By 10.6
mm SL (Fig. 197), larval Lampris have assumed the character-
istic adult, deep-bodied form. Larvae of the remaining genera
are also slender at hatching but become rapidly elongate with
growth. In these taxa, gut length may vary during ontogeny,
increasing to 80-90% SL in Lopholus. Eumecichthys and Re-
galecus. Gut length at transformation distinguishes these genera
from 7ai. Trachipterus, Dcsmodema and Radiicephalus.
Anterior dorsal rays are supported by a fleshy base in early
larval stages (Figs. 193, 195) and by modified radials first ap-
pearing as cartilage. The fleshy base and associated radials sup-
porting these rays are sometimes damaged in capture and torn
away from the cranium. This artifactual condition is occasion-
ally referenced in older literature as a "nuchal lobe"" (Hubbs,
1925:475) and "nuchal pennant" (Walters and Fitch, 1960:442-
443). In larval trachipterid, lophotid and regalecid fishes, the
first two dorsal elements are supported by elongate, stout radials,
the second of which serially supports succeeding radials of vary-
ing number (Table 100). These elements interdigitate in inter-
neural spaces and pterygiophore interdigitation patterns vary at
the generic level (Rosenblatt and Butler, 1977; Table 100). Law-
pris and Metavelifer possess a predorsal element which inter-
digitates in the first neural space. These genera (and presumably
Velifer) are unique in the possession of this character (Table
100). In addition, the morphology of the anterior modified ra-
dials varies with ontogeny. The rostrum or "horn" of Eume-
cichthys e\ongaies duringgrowth to twice its original length (Far-
Inlemeural space
Species
Z. cristatus
Trachipterus sp.
R. glesne
Desmodema spp.
R. elongatus
M. muitiradiatus
L. gultatus
1
7-9
1
2
1
8-9
1
2
1
11
2
2
7
T
1
13-14
1-2
1-2
P + 1
1
1
1
P + 1
1
1
1
in and Pokhil'skaya, 1968). Oelschlager (1976a) considers the
lophotid "horn'" to be supported by modifications of the frontals
and the supraoccipital (termed the "fronto-occipital carina"),
however these highly modified dorsal rays are likely supported
by dorsal fin radials which may fuse to cranial bones in adults.
In Dcsmodema. anterior dorsal rays are elongate in juveniles
but lost at transition (Rosenblatt and Butler, 1977). Presumably,
pterygiophores supporting these rays and interdigitating anterior
to the first neural spine (Table 100) are retained in adults.
Development of pelvic fin elements is precocious in all known
larval Lampriformes (Figs. 193, 195-200) and characterized by
reduction of ray number, length or both in some genera. In S.
chordalus (Figs. 199 and 200), a single elongate pelvic element
in early larvae becomes increasingly long and stout with de-
velopment, persisting until around 70 mm SL. Adult 5. chor-
datus (Regan, 1924; Table 98) lack a pelvic fin. A similar de-
velopmental sequence may occur in Desmodema which loses
pelvic elements by 173 mm SL (Rosenblatt and Butler, 1977)
and in R. elongatus (Harrison and Palmer, 1968). In Regalecus
and Agrostichthys {Oe\sch\a.ge:v, 1978a) the pelvic fin is retained
as a persistent larval floatation device that also serves in loco-
motion and taste perception.
The pectoral fin is the last to complete difTerentiation in larval
lampriformes. In our material, pectoral rays are incompletely
developed at 18.4 mm SL in R. elongatus (Fig. 198); 21.4 mm
SL in 5. chordatus (Fig. 200); and 29 mm SL in Trachipterus
sp. Pectoral development is most rapid in L. gultatus which
possesses adult counts by 10.6 mm SL (Fig. 197). Adults ofthis
species possess a strongly developed, lunate pectoral (Rosenblatt
and Johnson, 1976) which may also be important in locomotion
of larvae. Pectoral morphology and insertion vary consider-
ably among lampriform genera and are of systematic value,
although no comprehensive treatment exists.
Among lampriform genera, caudal morphology exhibits the
greatest potential for taxonomic and systematic evaluation. Ro-
senblatt and Butler (1977) have demonstrated the utility ofthis
character in distinguishing juvenile and adults of Trachipterus
and Desmodema and the details of caudal morphology (Table
98 and see Gosline, 1961; Hulley and Rau, 1969; Oelschlager,
1974; Patterson, 1968; Pietsch, 1978a; and Rosen, 1973 for
illustrations of caudal skeletons in various lampriform genera)
clearly delimit all other taxa, with the possible exception of
Agrostichthys {OehcMager, 1978b). Diff"erentiation of caudal
elements occurs early in development, rendering caudal mor-
phology an important larval identification criterion. Although
full developmental series are not available for most forms, the
OLNEY: LAMPRIFORMES
373
Fig. 195. Larvae of Zu cnstaius. (A) 6.5 mm NL and Regalecus glesne, (B) 5.4 mm; (C) 45.8 mm SL, all after Sparta 1933.
374
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 196. Larvae of Lampris gultalus. (Upper) 4.7 mm NL, MCZ 58990; (Lower) 8.6 mm SL, MCZ 58989.
OLNEY: LAMPRIFORMES
375
Fig. 197. Larva of Lampns gutlalus 10.6 mm SL. MCZ 58991.
caudal is complete (or nearly so) by 8.6 mm SL in L. guttatus
and 1 2-20 mm SL in R. elongatus. S. chordatus, Zu cristatus.
Trachipienis spp. and R. glesne.
The highly protrusible jaws of lampriform fishes develop pre-
cociously and jaw structures vary from the long, tubular mouth
of S. chordatus (¥\gs. 1 99 and 200; Pietsch 1978a) to only mod-
erately specialized in I'elifer. All lampriform genera possess
prcmaxillae with long ascending processes which fit withm the
nasal and ethmoid regions and slide forward during jaw pro-
traction. Larval lampriforms. especially Stylephorus (Figs. 199
and 200). are easily recognized by this feature although these
upper jaw specializations may not be unique to lampriform
genera (Rosen, 1973). In trachipterid, radiicephalid and regal-
ecid fishes, the premaxilla has a high, broad ascending process
which is often conspicuous in capture-damaged larvae.
Lampriforms are highly pigmented in all life history stages
376
ONTOGENY AND SYSTEMATICS OF nSHES-AHLSTROM SYMPOSIUM
Fig. 198. Larvae of Melavelifer multiradiatus. (Upper) 5.7 mm NL, MCZ 59717; and Radiicephalus elongatus. (Lower) 18.4 mm SL, ZMUC
uncatalogued. The vent is indicated by an arrow. The posterior portion of the ink gland is seen as a concentration of melanophores along the
ventral margin just posterior to the vent.
and larval pigment, especially in the form of melanophores
present laterally and along the dorsal and ventral margins, is
useful in identification of some genera (Figs. 193, 195-200).
Melanophores, concentrated on spatulate swellings in elongated
dorsal and pelvic rays help to distinguish larval lampriforms
although care should be taken since elongate, sometimes pig-
mented appendages are found in the larvae of a number of
unrelated taxa (Govoni et al., 1984). Among these taxa, how-
OLNEY: LAMPRIFORMES
377
Fig. 199. Larvae of Slylephorus chordatus. (Upper) 3.8 mm NL, MCZ 59718; (Lower) 7.6 mm SL, MCZ 59719.
ever, only lophiiform, bothid, zeid and serranid larvae have
elongate dorsal and pelvic elements.
Specialized ink glands filled with dark brown fluid are char-
acteristic of lophotid and radiicephalid fishes, and are conspic-
uous in larval R. elongatus (Fig. 198), young Lophotus (Fig.
201) and presumably Eumectchthys (Walters and Fitch, 1960;
Harrison and Palmer, 1968). Although this unpaired, internal
structure is not considered a larval specialization, its early ap-
pearance in larval R. elongatus and juvenile Lophotus suggests
that the ink gland may be functional in young fishes.
Development from larval to juvenile stages is gradual in Lam-
pris (Figs. 196 and 197; Oeschlager, 1976b) but ontogenetic
variability is marked and abrupt in trachipterid genera and Sty-
lephorus chordatus. This rapid transition from prejuvenile to
juvenile morphology has been termed metamorphosis in Des-
wof^ewfl (Rosenblatt and Butler, 1977) and Trachlptenis (Huhhs,
1925). In D. polystictum. metamorphosis is characterized by
changes in ventral profile, elongation of caudal vertebrae, in-
crease in eye size, eruption of mandibular teeth, and loss of
spots, pelvic fins and the posterior nostril (Rosenblatt and But-
ler, 1977). Examination of S. chordatus material indicates a
similar rapid transition, characterized by the loss of elongate
dorsal rays, three ventral caudal rays and stout pelvic elements
and by a marked change in eye morphology from a normal,
non-telescopic eye to the specialized adult condition (Pietsch,
1978a). Similar metamorphic change may occur in other lam-
priform taxa, however full developmental series are not avail-
able.
Meristics. — Mtn%X\c variability is useful in delimiting lampri-
form taxa (Tables 98, 100). Precaudal, caudal and total vertebral
counts distinguish genera and sometimes species (i.e., D. poly-
stictum vs D. lorum; T. fuku:aki vs T. altivelis) and total myo-
mere counts can be used to identify early larvae (Olney and
Naplin, 1980). Total vertebral/myomere counts of less than 53
characterize Lampris. Slylephorus and veliferids and are the
378
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 200. Larvae oi Stylephorus chordalus 21.4 mm SL, ZMUC uncataloged.
primary basis for the present identification of larval L. guttatus
(Figs. 196 and 197). Aletavelifer multiradialits {Fig. 198) and 5.
chordalus (Figs. 1 99 and 200). Total vertebral/myomere counts
range from 62-200 (Table 98) in trachipterid, radiicephalid,
lophotid and regalecid fishes, but care should be taken since
these elongate forms are often damaged in capture and the pos-
teriormost myomeres are difficult to discern in larvae.
Rays of median fins are either stout, unsegmented, spine-like
elements or typical soft rays. Previous researchers have been
inconsistent in their treatment of these elements as spines or
rays and ontogenetic variability likely exists. For these reasons,
total element counts are reported without reference to spine or
ray designation (Table 98). Furthermore, because of the lack of
developmental series in collections, few data exist on the se-
quence of development of these and other fin elements. As a
result, dorsal element counts delimit the genera of veliferid fishes
for example (Table 98), but are not developed in a 5.7 mm NL
Hawaiian specimen (Fig. 198). Identification of this specimen
(Fig, 198) as M. midtiradiatus is based on distributional records
(Walters, 1960; Heemstra, in press). In larval L. guttatus (Fig.
196; identification based on distributional records of Parin and
Kukuyev, 1983), total dorsal elements are developed by 8.6 mm
SL but counts indicate some overlap with veliferid species (Ta-
ble 98). In the elongate forms, dorsal element counts are less
valid identification criteria since complete differentiation of ele-
ments occurs late in development (approximately 44 mm SL in
Lophotus lacepedei, HML 6851; 83 mm SL in Eumecichthys
fiski. MCZ 42264).
The absence of the anal fin characterizes adult trachipterid
and regalecid fishes (Table 98), but its absence in early larvae
cannot be considered diagnostic. In genera possessing an anal
fin, differentiation of elements is evident in our material at 18
mm SL in R. elongatus (Fig. 198); 7.5 mm NL in S. chordatus
(Fig. 199); 33.5 mm SL in E.fiski; 5.7 mm NL in M. multiradia-
tus (Fig. 198); and 8.6 mm SL in L. guttatus (Fig. 196). Size at
first differentiation of anal elements of Lophotus spp. is un-
known but total element counts can serve to delimit young
Radiicephalus and Lophotus (Table 98; compare Figs. 198,201).
Larvae of these two forms can be easily confused due to the
common possession of the distinctive ink gland (Figs. 1 98, 20 1 ).
Total number of pectoral rays overlap considerably among
lampriform fishes and are of limited diagnostic value (Table
98). Total pelvic elements are of potential use in identification
but ontogenetic variability is great and care should be taken
until descriptions of full transformation series are available.
Total caudal elements are diagnostic among some lampriform
genera (Table 98), and, as previously discussed, details of caudal
morphology are important larval identification criteria.
Relationships
Our present knowledge of the egg and larval taxonomy of
lampriform fishes is inadequate to the task of fully understand-
ing phylogenetic relationships. Although larval stages have been
described for 8 of 1 2 genera (those of Agrostichthys. Desmo-
dema, Velifer and Eumecichthys remain unknown), full de-
velopmental series and detailed studies of developmental os-
teology and morphology are lacking. Among those taxa for which
some ontogenetic data are available, selected characters may
elucidate relationships within the Lampriformes and between
this group and other teleostean fishes. These are: (1) Egg mor-
phology. The distinctive eggs of lampriforms (Table 99, Figs.
193, 194) are likely specializations for epipelagic incubation
(Breder, 1962) and, if considered a derived condition, tend to
support the conclusion of a common ancestry for the group.
Complicating this interpretation is the lack of data on egg mor-
phology in all lampriform genera (Table 99) as well as the com-
mon possession of somewhat similar (although probably inde-
pendently evolved) egg morphology in other fishes (Orton,
1955a); (2) Precocious embryonic development. At hatching, all
known lampriform larvae possess fully developed, protrusible
jaws; functional, differentiated guts; and pigmented eyes. This
complement of precociously developed features shared by lam-
priform taxa may be a specialization for early, successful feeding
in the low prey densities of the epipelagic habitat. To my knowl-
edge, only exocoetoid fishes exhibit similar development; (3)
Elongate anterior dorsal elements. All known lampriform larvae
OLNEY: LAMPRIFORMES
379
Fig. 20 1 . Photomicrograph of the posteriormost ponion of the ink
gland in young Lophotus lacepedei (HML 6851, 45 mm SL). Ink gland
is seen as the dark, tubular body overlymg the hindgut and vent. The
vent is indicated by an arrow.
possess elongate anterior dorsal elements which are ornamented
with spatulate, pigmented swellings in some genera. As with
other fishes with ornamented larval appendages (Govoni et al.,
1984), variation in ornamentation may be due to capture dam-
age. As a result, the absence of elaborate ornamentation in early
larvae of L. guttatus (Fig. 196), M. imdtiradiatus (Fig. 198) and
S. chordatus (Figs. 199 and 200) could be artifactual; (4) Pelvic
fin elements. Precocious appearance of ventral fin elements which
are stout, elongate and supported by well developed pelvic bones
is observed in all known lampriform larvae (Figs. 193, 195-
200). Variation among genera occurs in element number and
fate at metamorphosis. In I'elifer and Lampris, pelvic elements
are numerous and well developed in adults. In remaining genera,
reductive trends are evident and only regalecids retain strongly
developed and specialized pelvic fins (Oelschlager, 1978a); (5)
Minute spines on dorsal elements. Small, laterally projecting
spines are conspicuous in some young lampriform fishes and
have been reported in juveniles by Walters and Fitch (1960:
443), Rosenblatt and Butler ( 1 977:844), and Heemstra and Ken-
nemeyer (in press). In our material, these minute spines are
conspicuous in larval Zu. Trachiptenis. Regalecus. Lophotus
and Radiicephalus as well as juvenile specimens of Desmodema
and Eumecichthys. Larval Lampris. Metavelifer and Stylephonis
lack these characters; (6) Multiple pterygiophores interdigitate
in first two interneural spaces. In all our lampriform material,
only L. guttatus and M. multiradiatus have fewer than seven
pterygiophores which interdigitate in interneural spaces 1 and
2 (Table 100). In addition, only Lampris and Metavelifer (and
presumably Velifer) possess a single predorsal element. Inter-
digitation sequences in Velifer. Lophotus. Eumecichthys. Sty-
lephorus and Agrostichthys are unknown; and (7) Metamorpho-
sis. The absence of abrupt ontogenetic transition delimits Lampris
(and presumably veliferids) from other lampriform genera.
The distribution of ontogenetic characters 1-7 among lam-
priform genera may be instructive when considering suggestions
by previous authors of evolutionary trends within the order.
The indication of monophyly by Regan (1907, 1924) and the
adoption of this hypothesis by Greenwood et al. (1966) and
Oelschlager (1976a) is supported by the common possession of
characters 1-4 among all known lampriform larvae. Rigorous
testing of this hypothesis utilizing ontogenetic data, however,
must await a more complete knowledge of egg and larval de-
velopment among Lampriformes and between these fishes and
other groups. Rosen (1973) suggested that relationships among
trachichthyoids, berycids, zeoids and lampriforms seem plau-
sible. Ontogenic characters (1-4) which appear to unite the di-
verse lampriform genera are variously present, absent or un-
known in trachichthyoid, berycid and zeoid fishes and present
no clear picture of inter-relationships. Larvae of Diretmus and
Diretmoides (Post and Quero, 1981) lack these characters and
are distinguished by pronounced occipital and preopercular
spines. Polymixia sp. ( 1 0.0 mm SL; MCZ 58964) lack characters
2 and 3 (eggs of Polymixia are unknown) but possess well
developed ventral fins. These fins may not be present at hatch-
ing, however. Juvenile Cyttus traversi (James, 1976b) possess
elongate, ornamented and pigmented pelvic and anterior dorsal
elements, although the sequence of development of these struc-
tures is unknown. The rhomboidal body shape, symmetrical
caudal and jaw structure of C. traversi resemble deep-bodied
lampriform genera.
Rosen and Patterson (1969), Rosen (1973) and Oelschlager
(1974, 1976a, 1978a, b, 1979) have examined osteological and
functional aspects of adult lampriform morphology and com-
mented on relationships. Recent fishes are represented by a
series of highly modified forms of which i'elifer is believed to
be the least specialized. Veliferids are considered to be more
closely related to Lampris than to any other genus on the basis
of similar body form, caudal morphology, meristics and the
possession of a predorsal element. No apomorphous character
serves as a criterion for monophyly in the Veliferidae (Oel-
schlager, 1976a). Ontogenetic characters 5-7, however, may be
useful in defining relationships between the two series of families
[Oelschlager's (1976a) Bathysomi and Taeniosomi] within the
order.
Among the elongate genera, Agrostichthys is considered most
closely related to Regalecus (Oelschlager, 1978a, b). Desmo-
dema and Zu represent an apomorphous sister group of Tra-
chiptenis. considered the most primati ve trachipterid genus (Ro-
senblatt and Butler, 1977). Radiicephalus appears to be the least
specialized among all elongate lampriforms although it shares
several specialized features (ink sac, caudal filament) in common
with lophotids and Stylephorus (Harrison and Palmer, 1968).
Virginia Institute of Marine Science, College of William
AND Mary, Gloucester Point, Virginia 23062.
Mirapinnatoidei: Development and Relationships
E. Bertelsen and N. B. Marshall
FISHES of the Mirapinnatoidei are soft-rayed, scaleless,
oceanic teleosts with elongated body, jugular pelvic fins of
4-10 rays, a single dorsal fin opposed to the anal fin with origin
behind mid-standard length, pectorals lateral, caudal fin with
10 -(- 9 principal rays, cleft of mouth oblique to subvertical,
premaxillae excluding maxillae from gape, jaws no more than
slightly protrusible, branchiostegal rays 3-5 on epihyal, 4 on
ceratohyal, swimbladder, functional only in larvae, with two
posterior retia mirabilia that supply an anterior gas gland. An
isolated phylogenetic lineage of uncertain systematic position
but apparently most closely related to the Megalomycteroidei
and the Cetomimoidei.
Development
These fishes were originally placed in two families by Ber-
telsen and Marshall ( 1 956): ( 1 ) Mirapinnidae, with a single genus
and species Mirapinna esau(Fig. 202) based on a single subadult
female 39.5 mm SL caught at the surface off the Azores and (2)
Eutaeniophoridae, with two genera Eutaeniophorus and Para-
taeniophorus (Figs. 202 and 203) both known only in larval and
metamorphosis stages less than 55 mm SL that are epipelagic
in tropical and subtropical parts of all oceans.
Examination of more recent material indicates that these fish-
es are better regarded as members of a single family. Mirapin-
nidae, containing the above mentioned 3 genera. Adults prob-
ably are mesopelagic. The genera and species are distinguished
by meristic and morphometric characters as well as differently
developed dermal structures (Table 101). Hair-like outgrowths
of the epidermis are found all over the head, body and fins of
Mirapinna esaii. The longest hairs measure from about 1 .0 to
1.5 mm in length and bear stalked glandular cells. The skin of
Eutaeniophorus and Parataeniophorus is densely covered with
minute papillae less than about 0.05 mm in length (Bertelsen
and Marshall, in preparation). Skin of the caudal fin of Eutae-
niophorus and Parataeniophorus is prolonged into a ribbon-like
streamer reaching lengths of 200-300% SL. Upper and lower
lobes of the caudal fin overlap in Mirapinna.
Specimens. — \nc\ud\ng a number of unpublished records the
material of Mirapinnatoidei known to us consists of: One Mir-
apinna esau: the holotype, a 39.5 mm juvenile female; about
100 Eutaeniophorus festivus 8.0-53 mm; two Eutaeniophorus
n. sp. (in preparation) 12 and 16 mm; 32 Parataeniophorus
gulosus 8-35 mm; 3 Parataeniophorus hrevis 13.5, 29 and 46
mm; 2 Parataeniophorus n. sp. (in preparation) 9 and 1 1 mm;
about 40 unidentified small larvae (most probably E. festivus)
5-12 mm. Eggs of Mirapinnatoidei are unknown and no larval
Mirapinna has been recorded; [according to our reexamination
a specimen of about 16 mm referred to this species by Four-
manoir, (1971b) is a Parataeniophorus sp.]. All the specimens
have small immature gonads. A light brown pigmentation of
the skin appears at a larval length of about 20 mm and some
of the 35-53 mm largest specimens are dark brown and are
considered post-metamorphic juveniles. However the transfor-
mation from larval to juvenile appearance is quite gradual with-
out any distinct specialized metamorphic stage.
The youngest Eutaeniophorus larva described (6.5 mm SL)
has remains of a yolk sac, nearly unpigmented eyes, no rudiment
of pelvic fins, continuous embryonic fins without trace of fin
rays and, except for a ventral series of melanophores, the body
is completely unpigmented (Bertelsen and Marshall, 1958). Full
numbers of rays of the unpaired fins may be delected at 8-9
mm SL. Rudiments of pelvic fins are present at 6-7 mm SL,
the number of rays discemable at about 10 mm SL. Pectoral
fin rays are not well differentiated until lengths of about 20 mm
SL. The caudal streamer, characteristic of Eutaeniophorus and
Parataeniophorus, is present as a short rudiment in the 5-6 mm
youngest larvae; it increases with increasing SL. It is broken in
most specimens of more than about 10 mm. The greatest lengths
observed are 86% SL in a Parataeniophorus hrevis of 22.6 mm,
about 200% SL in two specimens of Eutaeniophorus of 12-15.5
mm SL (unpublished data), and no less than 300% in an E.
festivus of 35 mm (Fig. 203).
All Eutaeniophorus larvae are very slender with body depth
less than 10% SL except for the largest specimens. Body depth
in Parataeniophorus species is less than 1 5% SL. Predorsal lengths
(snout to first dorsal finray) in these genera is 69 to 77% SL (cf
Table 101),
All larvae have a fine peppering of melanophores on head
and body, slightly increasing in density with increasing SL, with
no distinct grouping except for a slight increase in density to-
wards the tail, on the dorsal part of the peritoneum, and along
Table 101. Characteristics of Mirapinnatoidei.
Eutaenio-
Mirapinna
esau
Parataeniophorus
hrevis gulosus
phorus
festivus
Texture of skin
Hairs
Minute papillae
Caudal streamer
Absent
Present
Ungth in % SL:
Head
27
10-17
12
10-13
Predorsal
61
71-7:
69
10-11
Longest pectoral finray
14
3-9
7
4-7
Longest pelvic finray
40
15-25
16-39'
12-19^
No. of finrays:
Pectoral
13
ca. 16-18
19-20
20-24
Pelvic
8
8-9
9-10
4-5
Dorsal
16
15-20
28-33
16-20
Anal
14
14-17
23-29
15-18
No. of vertebrae:
Predorsal
22
21-27
28
31-36
Under dorsal fin
10
8-11
13
8-11
In caudal peduncle
16
10
8
7-8
Total
48
42-46
49
47-55
' Shorter in specimens less than 15 mm.
- Shorter in specimens less than 25 mm.
380
BERTELSEN AND MARSHALL: MIRAPINNATOIDEI
381
■■■ \ I / / /
«a^
10 mm il ! M ^ \ ,\ V
""'^-.
Fig. 202. (A) Mirapinna esau. holotype. 39.5 mm SL; (B) Paralaeniophorus brevis. 29 mm SL; and (C) Parataemophorus brevis. holotype,
13.5 mm SL. A and C from Bertelsen and Marshall (1956), B drawn by Kai L. Elsman.
the myosepta. Density of pigment is greater on the caudal
streamer and the caudal fin rays at the base, the fully developed
streamer has a median longitudinal band of pigment and a nearly
black ventral border. The two species of Parataemophorus differ
in pigmentation from Emaeniaphorus feslivus in having a distal
patch of pigment on each pelvic fin. No other distinguishing
characters in pigmentation have been found.
Relationship.s
Reference to Bertelsen and Marshall ( 1 956), Myers and Frei-
hofer (1966) and Goodyear (1970) shows that both the mira-
pinnatoid and megalomycteroid fishes have the following com-
mon features; ( 1 ) they are small, elongated fishes with a relatively
small head and mouth; (2) the suspensoria are inclined forwards
and there is a single supramaxilla in the upper jaw; (3) they have
382
A
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
'„CLI ■3(.W."1 Ij'A
10 mm
Fig. 203. (A) Eutaeniophorus festivus. paratype, 35 mm SL, with complete caudal streamer, 106 mm in length; (B) Eutaeniophorus festivus.
holotype. 53 mm SL; and (C) Paralaeniophorus gulosus. paratype, 21 mm SL. All from Bertelsen and Marshall (1956).
soft rays, which are unbranched [except in the caudal fin of
megalomycteroids? Myers and Freihofer's (1966) drawing of
Megalomycter leevani shows the complete caudal rays ending
in actinotrichia, and they state that the dorsal and anal rays are
unsegmented]; (4) the dorsal and anal fins are opposed and
inserted on the posterior half of the body; (5) the pectorals are
laterally set and have numerous rays (D. 15-33, A. 14-29 in the
mirapinnatoids; D. 15-26, A. 14-20 in the megalomycteroids);
(6) the pelvic fins are inserted below or before the base of the
pectorals, but are reduced or absent in the megalomycteroids,
whereas the pelvics of the mirapinnatoids are well developed;
(7) the numbers of branchiostegal rays (on the epihyal and cer-
atohyal) are 3-5 + 4-5; (8) the vertebrae number 41-54 (45-
48 in the megalomycteroids).
The main differences between the two groups concern the skin
(papillate or "hairy" in the mirapinnatoids, scaled in the me-
galomycteroids), olfactory organs (very large in the latter, small
in the former) [Goodyear's (1970) specimen was a ripe male but
Myers and Freihofer (1966) did not determine the sex of their
specimens. It is possible that the females have yet to be found
and are microsmatic]. The gape markedly oblique in the mir-
apinnatoids, somewhat oblique or horizontal in the megalo-
mycteroids.
The mirapinnatoids resemble the cetomimoids in having soft
rays, a scaleless skin, opposed dorsal and anal rays on the pos-
terior part of the body and the same numbers and arrangement
of branchiostegal rays (mirapinnatoids 3-5 + 4, cetomimoids
3-4 + 4-5). There is also a marked resemblance between the
swimbladder of Barbourisia. which regresses after a presumed
functional stage in the early life history, and the swimbladder
of the mirapmnatoids (see Bertelsen and Marshall, 1956). In
both there are two posterior retia mirabilia that run forward to
an anterior gas gland.
One main difference between these two suborders concerns
the head, which whether relatively large or small in the ceto-
mimoids, bears long jaws with a more or less horizontal gape.
This contrasts strongly with the relatively short, upturned jaws
of the mirapinnatoids. (Even so, it may well be that the fishes
of these suborders and the megalomycteroids feed largely on
copepods.) Secondly, in the two cetomimoids that have pelvic
fins {Rondcletia and Barbourisia) these are abdominal in posi-
tion whereas those of the mirapinnatoids are jugal.
Beside the similarities considered above, the mirapinnatoids,
megalomycteroids and cetomimoids resemble each other in the
disposition of the red muscle component of their axial muscles.
Down the entire length of their myotomes red muscle fibres
BERTELSEN AND MARSHALL: MIRAPINNATOIDEI
383
cover at least the main "V" of each element, and such an ar-
rangement seems to be unusual in teleosts. Similar wide red
muscle coverage of the myotomes is found also in the stomia-
toids and giganturoids (Marshall, unpublished) and apparently
alsoinmaleceratioidangler-fishes(Marshall, 1971).Othergroups
will probably prove to have this kind of red muscle arrangement
but the most usual condition in teleosts is a narrow concentra-
tion of red muscle on either side of the horizontal septum down
the entire length of the fish. However, in alepisauroids the ver-
tical extent of red muscle expands towards the tail, where it may
cover most of the myotomes (Marshall, 1971; Johnson, 1982).
The above treatment of adult characters indicates that the
mirapinnatoids are most closely related to the megalomycter-
oids. Next to the latter they are most nearly allied to the ceto-
mimoids. As will be seen from the title of this paper, we have
followed Greenwood et al. (1966) in placing all three suborders
in the order Cetomimiformes away from the Acanthopterygii.
Whether they can be gathered into a larger ordinal grouping, as
in the Lampridiformes (Rosen and Patterson, 1969) or in the
Beryciformes (Rosen, 1 973), is a matter for further comparative
studies (see also Zehren, 1979). Nothing is known of larval
megalomycteroids and cetomimoids. Larval forms of other
groups seem to have no affinities to larval mirapinnatoids.
(E.B.) Zoological Museum, University of Copenhagen,
Copenhagen 2100 0. Denmark. (N.B.M.) 6 Park Lane,
Saffron-Walden. Essex, England.
Beryciformes: Development and Relationships
M. J. Keene and K. a. Tighe
IN the classification of Greenwood el al. (1966), the order
Beryciformes was divided into 3 suborders; the Stephan-
oberycoidei with 3 families, the Polymixoidei with I family and
the Berycoidei with 8 families. Rosen and Patterson (1969) re-
moved the Polymixiidae from the Beryciformes, assigned it to
a new order, the Polymixiiformes and placed this order in the
Paracanthopterygii. Rosen and Patterson (1969) also moved the
Cetomimidae, Barbourisiidae and Rondeletiidae to the Bery-
ciformes in the suborder Cetomimoidei. Woods and Sonoda
(1973) considered the order Berycomorphi to contain the fam-
ilies Polymixiidae, Diretmidae, Monocentridae, Anomalopidae,
Trachichthyidae, Holocentridae, Berycidae, Sorosichthyidae, and
Table 1 02. Merlstic Ranges, OsTEOLooirAL Characters, Number of Genera,
All data are from Woods and Sonoda (1973), Ebeling and
AND Number of Species for Families in the Order Beryciformes.
Weed (1973) or Zehren (1979) unless noted.
Osteologjcal characters
Principle
. Number
Number
caudal
Branchi-
Verte-
Orbito-
Subocular
Supra-
of
of
Pelvic
Dorsal
Anal
rays
Pectoral
ostegals
brae
sphenoid
shelf
maxillarv
genera
species
Berycidae
I, 7 or
I, 10-12
iv-vn,
13-18
IV,
12-30
16-17
13-18
7-9
24
Present
Present
2
2
ca. 10
Anoplogasteridae
1,6
0, 17-19
0, 8-9
17
14-16
8
28
Present
Present
1
1
1
Diretmidae
1,6
0, 17-19
0, 18-24
17
16-20
7-9
27-31
Present
Present
1
2
3
Soroshicthyidae'
1,5
X, 8
11.8
16 or 17
13
7
?
9
7
7
1
1
Trachichthyidae
1,6
IV-VIII,
12-18
II-IIl,
8-12
17
14-20
8
26-29
Present
Present
2
5
ca. 14
Anomalopidae
0, 6-7
n-iv,
14-19
II, 10-13
19
16
7 or 8
25-30
Present
Present
1 or 2
3
4
Monocentridae
I, 2-3
IV-VII,
11-12
IV-V,10
0. 10-12
19
14
8
27
Present
Present
1
1
2
Hispidoberycidae-
1, 6
III, 9
17
12
8
34
Present
Absent
1
1
1
Holocentridae
I. 5-8
X-XIII,
11-16
IV, 9-16
17
14-17
7-8
26-27
Present
Present
2
9
ca. 70
Gibberichthyidae
0,5-6
V-VIII,
8-9
IV-V,
7-9
17
13-15
8
28-31
Absent
Absent
1
1
2
Stephanoberycidae
0,5
O-III.
11-14
O-III,
10-13
19
11-18
7-8
30-33
■Absent
Absent
1
3
3
Melamphaidae
I, 6-9
I-III,
9-18
I, 7-10
17
14-17
24-31
Absent
Absent
Oor 1
5
ca. 30
Rondeletiidae
0, 5-6
0, 13-16
0, 13-15
17
9-10
8
24-27
Absent
Absent
1
1
2
Barbourisiidae
0. 6
0,21-22
0, 16-18
17
14
8
ca. 42
Absent?
Absent?
1
1
1
Cetomimidae
Absent
14-ca. 30
13-ca. 30
11-19
16-20
8-10
51-52
Absent?
Absent?
1
5
10
Paradirctmidae'
1,5
X-XI?,
15?
III, 13
ca. 13
16
7
9
7
7
7
1
1
' Whitley, l'>45.
'Kotlyar, 1981,
'Whitley. 1946,
384
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 103. References Giving Descriptions and/or Figures of
Early Life History Stages of the Order Beryciformes.
Pre-juveniles/
Family
Eggs Larvae
juveniles
Holocentridae
- McKenney, 1959
McKenney, 1959
Jones and
Jones and
Kumaran, 1962
Kumaran, 1962
Aboussouan, 1966b
Greenfield, 1965
Randall et al.,
1982
Melamphaidae
- Ebeling, 1962
Ebeling and Weed,
1963, 1973
Pcrtseva-Ostrou-
mova and Rass,
1973
Moser, unpub-
lished
Diretmidae
- Post, 1976
Post and Quero,
1981
Anoplogasteridae
— —
Parr, 1933
Trachichthyidae
— Ahlstrom (notes)
Parr, 1933
Crossland, 1981
Johnson, 1970
Gibberichthyidae
Robins and
de Sylva, 1965
Thorp, 1969
de Sylva and
Eschmeyer.
1977
Berycidae
Anomalopidae
Monocentridae
Sorosichthyidae
Paradiretmidae
H ispidoberycidae
Stephanoberycidae
Barbourisiidae
Rondeletiidae
Cetomimidae
Anoplogasteridae. while Ebeling and Weed (1973) considered
the order Xenoberyces to contain the families Melamphaidae,
Gibbeinchthyidae, and Stephanoberycidae. Both pairs of au-
thors gave diagnostic characters, and compared and contrasted
their orders. Zehren (1979), after studying the comparative os-
teology and phylogeny of the beryciform families of Greenwood
et al. ( 1 966), also concluded that the Polymixiidae did not belong
in the Beryciformes. Nelson (1976) included the families So-
rosichthyidae and Paradiretmidae in the suborder Berycoidei
but did not treat them further. Kotlyar (1981 ) described a new
species of beryciform which he felt deserved status as a new
family, the Hispidoberycidae. He tentatively aligned his new
family within the Berycoidei. The Beryciformes are presently
defined on the basis of several primitive characters such as the
presence of an orbitosphenoid and subocular shelf (in most
forms) and a high number of pelvic and caudal rays as well as
several derived characters such as the presence of dorsal, anal
and pelvic spmes. and the presence of spinous procurrent caudal
fin rays. However, none of the characters is unique to the order
and the monophyly of the order is still in question. Meristics,
osteological characters, and the number of genera and species
in each beryciform family are shown in Table 102.
Although the systematics of the Acanthopterygii is in a state
of flux, the order Beryciformes presently contains 3 suborders;
the Stephanoberycoidei with 3 families, the Berycoidei with 10
families, and the Cetomimoidei with 3 families.
The Beryciformes are considered by Greenwood et al. (1966)
to be the basal stock from which some of the more advanced
acanthopterygians have evolved. Beryciformes are marine and
occur in all oceans. Some species are semibenthic inhabiting
coral reefs, rocky shores, and shelf or slope waters (Woods and
Sonoda, 1973) while others are epipelagic, mesopelagic. bathy-
pelagic, or bathybenthic (Ebeling and Weed, 1973).
Development
There is no published information on early life history stages
for the Monocentridae, Anomalopidae, Berycidae, Sorosichthy-
idae, Paradiretmidae, Hispidoberycidae, Stephanoberycidae,
Barbourisiidae, Rondeletiidae, and Cetomimidae (Table 103).
Although information is lacking on the eggs of the Beryciformes,
there is some on other early life history stages of the Holocen-
tridae, Melamphaidae, Anoplogasteridae, Diretmidae, Trach-
ichthyidae and Gibberichthyidae.
The Holocentridae contains two subfamilies, the Holocen-
trinae and the Myripristinae. Prejuveniles and early life history
stage series are known for at least one species in each subfamily.
McKenney (1959) gave a detailed description of the early life
history of Holoccntrus ve.xi/lanus based on specimens less than
2.0 mm to adults, while both Aboussouan (1966b) and Jones
and Kumaran (1962) figure and discuss larvae of Holocentrus
sp. less than 5.0 mm SL. Jones and Kumaran (1962) also figure
and describe larval stages ranging from 2.7 to 6.7 mm, for My-
ripristis mirdjan [specific identification questioned by Greenfield
(1965)]. McKenney (1959) figured the prejuvenile or rhynch-
ichthys stage of Holocentrus vexillarius while Jones and Ku-
maran ( 1 962), Greenfield (1965), and Randall et al. (1982) figure
the rhynchichthys stage for several myripristine species. The
following characterization of holocentrid development is based
on the data of McKenney (1959) and Jones and Kumaran (1962).
Holocentrid larvae are characterized by a relatively large head
with well-developed preopercular, rostral, and median cranial
spines. Pigmentation is extensive on the peritoneum and there
is a ventral line of melanophores in the postanal region. The
long preopercular spines develop first and are well developed
at 1.8 mm TL (Fig. 204A). At 2.2 mm the posteriorly directed
cranial spine is rapidly forming and by 2.8 mm the rostral spine
is apparent. The 5.0 mm H, vexillarius and 4.7 mm Mynphstis
sp. (Fig. 204B, C) both exhibit strong rostral, median cranial,
preopercular and opercular spination that develops into the head
armor found in the rhynchichthys stage (Fig. 204D, E). There
Fig. 204. (A) Preflexion larva oi Holocentrus vexillarius. 1.8 mm NL (source: McKenney, 1959); (B) Flexion larva o[ Holocentrus vexillarius.
5.0 mm NL (source; McKenney, 1959); (C) Preflexion larva of A/ir/pmrKsp.. 4.7 mm NL (source; Jones and Kumaran, 1962); (D) Rhynchichthys
prejuvenile of Holocentrus vexillarius. 24.9 mm SL (source; McKenney, 1959); and (E) Rhynchichthys prejuvenile of Myriprislis sp., 16.3 mm
SL (source: Jones and Kumaran, 1962).
KEENE AND TIGHE: BERYCIFORMES
385
386
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 205. Urval senes oi Melamphaes lugubris (A) 5.3 mm NL, (B) 6.2 mm SL. (C) 10.4 mm SL and (D) 15.2 mm SL (source: Southwest
Fisheries Center, CalCOFI, original, illustrated by B. Y. Sumida).
KEENE AND TIGHE: BERYCIFORMES
387
Fig. 206. (A) Postflexion larva of Scopelogadus bispinosus. 8.0 mm SL; (B) Postflexion larva of Poromitra sp.. 13.5 mm SL; (C) Postflexion
larva of Poromitra megalops. 10.0 mm SL; (D) Postflexion larva of Scopelobery.x sp., 6.5 mm SL; (E) Postflexion larva of Melamphaes lugubris,
8.3 mm SL; (F) Postflexion larva of Melamphaes lyphlops. 9.4 mm SL (source: Ebeling, 1962).
are minor differences in the spine patterns of the two species
illustrated, and the Holocenlnis spination is somewhat more
developed. All of these spines are lost as the fish develops into
a juvenile in the Myripristinae, while the Holocentrinae retain
only large preopercular spines.
A considerable amount is known about at least some of the
early life history stages of all melamphaid genera except Sio.
Notes from Moser and Ahlstrom, and an examination of me-
lamphaid larval specimens from the Southwest Fisheries Cen-
ter (SWFC), allow the following conclusions to be made about
early melamphaid larvae 2-10 mm for Poromitra. Melam-
phaes. Scopelogadiis, and Scopeloberyx. Specimens in this range
tend to have a relatively more elongate and slender body shape
than later larval stages. The pelvics develop rapidly followed
closely by the pectoral fins. The pelvic fin origin is more anterior
than in later stages, and the pelvic rays are quite long, fragile
and darkly pigmented. This condition persists longer and is
more striking in some species such as M. lyphlops than in others.
In early larvae of Melamphaes. Scopelogadiis. and Scopelobe-
ryx. two pigment spots occur near the posterior end of the dorsal
and anal fin anlagen (Fig. 205A). These pigment spots spread
both anteriorly and postenorly during growth to form longitu-
dinal rows of pigment along the dorsal and ventral surfaces of
the body (Fig. 205B). In some species, these areas of initial
pigmentation spread laterally to form a band of pigment between
the dorsal and anal fin bases in later larval stages. Additional
pigmentation occurs on the cranium and peritoneum in all four
genera, and in the form of a spot at the posterior end of the
caudal peduncle in at least Melamphaes and some Poromitra.
In these early stages, the second or third dorsal fin ray tends to
be much longer than the others, extending to the region of the
caudal peduncle. This elongate ray is known to occur in Me-
lamphaes. Scopeloberyx. and Scopelogadus. Usually damaged.
this elongate ray is not evident after 5-10 mm but, even in
adults, the second or third dorsal ray (spine) is the largest. By
5-10 to 20 mm SL melamphaid larvae exhibit body shapes
and other characters such as meristics and preopercular spi-
nation that allow them to be separated into genera (Ebeling,
1962; Fig. 206A-F). Development is gradual and direct; there
are no known prejuvenile stages. Additional larvae were illus-
trated and are published here without further comment (Fig.
207).
Early life history stages are known for all three species con-
tained in the two genera of the Diretmidae. Post (1976) discusses
the systematics and early life history of two of these species,
and Post and Quero (1981) in their familial revision, describe
a new genus and species, give the early life history of all three
species, and provide a key for the identification of juveniles.
The larvae of all three species are relatively elongate at 4-5 mm
sizes but rapidly develop a relatively deeper body. All three
species also possess a short stout spine over each eye, a longer
cranial spine directed posteriodorsad on each side of the head,
and a long preopercular spine directed posterioventrad (Fig.
208A). The head spine configuration is quite similar to that of
.4. cormita. described below, and is gradually lost during growth.
The monotypic Anoplogasteridae contains the highly spe-
cialized mesopelagic predator .-inoplogaster cormtta. Specimens
over about 100 mm SL are jet black with large fangs while
specimens less than 80 mm SL are metallic grey with black
pigmentation developing along the ventral midline, do not have
such large teeth, and exhibit a pattern of head spination not
found in larger individuals (Woods and Sonoda, 1973). USNM
collections contain many individuals from 4.5 mm TL larvae
to adults, upon which the following characterization of the early
life history stages is based.
A 4.5 mm prefiexion larva has the caudal fin elements de-
KEENE AND TIGHE: BERYCIFORMES
389
Fig. 207. (A) Preflexion larva of Scopeloberyxsp.. 4.4 mm XL; (B) Postflexion larva of Scopeloberyx opisthopterus. 9.1 mm SL; (C) Postflexion
larva of Scopeloberyx robuslus. 1 3.0 mm SL; (D) Preflexion larva of Poromitra crassiceps complex, 7.9 mm SL; (E) Postflexion larva of Melamphaes
lepnis. 19.5 mm SL; all drawn by B. Washington.
veioping. The dorsal, anal and pectoral fins are already devel-
oped, while pelvic fin buds are present. The pattern of head
spination described below is already well fisrmed. A 6.0 mm
postflexion specimen (Fig. 208B) has all fins completely devel-
oped except for the pelvics and procurrent caudal elements.
There is pigmentation on the head, lateral surface of the body
and caudal peduncle, while the abdominal area is pale with
scattered melanophores. A small pigmented area occurs on the
pectoral bases. A serrate frontal ridge bordenng the anterior of
each eye terminates in a short stout supraocular spine. Ridges
continuing posteriodorsad on the cranium terminate in long
serrate spines probably arising from the parietals. The pre-
opercles end in strong serrate spines directed posterioventrad.
By 9.0 mm SL, the pelvics have become well-developed and
the head spination is still strong. A small dense patch of me-
lanophores occurs on the ventral surface of the body justantenor
to the origin of the pelvic fins. With increasing growth (28 mm
SL), this pigmentation darkens and expands, extending forward
in a continuous band to the tip of the isthmus. Additional pig-
mentation occurs at the ongin of the pelvic fins, around the
vent, just posterior to the anal fin on the caudal peduncle, and
in a transverse bar on the abdomen midway between the pelvic
origin and the vent. The increase in dark pigmentation and the
reduction in cranial and preopercular spines in larger juveniles
is described by Woods and Sonoda (1973).
Crossland (1981) illustrated a trachichthyid larva, probably
of OpliYus I'longatus. taken off northeastern New Zealand (Fig.
208C). Larger larvae of the same species had the skin on the
dorsal surface of the head and body covered with tiny spines.
Ahlstrom (notes) sketched early Trachichthys mento larvae that
are fairly deep-bodied at 3.5 to 4.5 mm, with the pectoral fin
showing precocious development. A dark spiny pigmented band
extending from the region of the anal to the dorsal occurs in 3.5
mm TL specimens. This spination covers areas on both sides
of the dorsal fin, parts of the head, thoracic region and jaws. In
a preflexion 6.4 mm specimen, the fin rays are mostly developed,
and the body is stockier, approaching the shape of the adult and
is covered with minute spines. The holotype of Korogaster nanus
Parr 1933, synonomized by Woods and Sonoda (1973) in Ho-
plostethus. is 19 mm long (Fig. 208D), possesses unbranched
rays in the pectoral, pelvic and caudal fins, and has dermal
papillae and small spines all over its body. This specimen and
the second specimen of Korsogaster Tepone:<i by Johnson (1970)
(Fig. 208E) are juveniles of the family Trachichthyidae.
The most striking early life history of any beryciform is ex-
hibited by the prejuvenile kasidoron stage of gibberichthyids
(Figs. 209, 210). This stage is characterized by a long trailing
pelvic appendage which is part of a modified third pelvic ray
and is present in specimens from at least 7.5 to 21 mm TL. It
is lost by 30 mm SL (de Sylva and Eschmeyer, 1977). Dunng
early growth, this trailing appendage becomes more ornate and
resembles the trailing tentacles of siphonophores or Sargassum
weed at about 15 mm SL. Up until about at least 20 mm, the
prejuveniles inhabit epipelagic waters but by 30 mm individuals
have lost the pelvic appendage and taken up a mesopelagic to
upper bathypelagic existence. The anterior dorsal and anal fin
elements are soft rays during the kasidoron stage, but develop
into strong fin spines in the adult. There is also a marked de-
velopment of bony head ridges in the adults, that is not found
in the stages 20 mm and smaller (de Sylva and Eschmeyer,
1977).
Relationships
Rosen and Patterson (1969) and Rosen (1973) emphasized
the futility of the present classification of the Beryciformes and
the rest of the Acanthopterygii, because it relies on grouping of
primitive characters to express relationships. Realizing this.
Zehren (1979) did a phylogenetic analysis of the Beryciformes
to attempt to determine whether or not the order is monophy-
letic(Fig. 211).
Besides supporting Rosen and Patterson's removal of the
Polymixiidae from the Beryciformes, Zehren's analysis super-
ficially suggests that the remaining ten families form a mono-
phyletic group. However, he cautions that since none of the
derived character states that he uses is unique to the ten families,
their monophyly is uncertain.
Zehren's results and discussion suggest that the Holocentridae
do not appear to be closely related to the other nine families
studied. Woods and Sonoda (1973) felt that the Holocentridae
were very different from the other Beryciformes and Rosen (1973)
390
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 208. (A) Postflexion larva of Direlmus argenteus. ca. 6 mm SL (source: Post, 1976); (B) Postflexion larva oi Anoplogaster cornuta. 6.0
mm SL (USNM 244902) drawn by B. Washington; (C) Flexion larva of Opiivusus elongatus ? 5.3 mm NL (source: Crossland, 1981); (D)
Unidentified trachichthyid juvenile, 19.0 mm SL (source: Parr, 1933); (E) Unidentified trachichthyid juvenile, 21.5 mm SL (source: Johnson,
1970).
KEENE AND TIGHE: BERYCIFORMES
391
Fig. 209. Kasidoron prejuvenile of Gibberichthys pumilus, 1 5.3 mm SL (source: de Sylva and Eschmeyer, 1977).
considered the holocentrids to be a distinct major subgroup
within the order. Rosen (pers. comm. to Zehren) believes that
the Holocentridae should be placed within the Perciformes.
Another result of Zehren's study is that the Berycidae appear
to be the primitive sister group to the other eight families and
should be placed in their own suborder, the Berycoidei. The
Trachichthyidae, Diretmidae. Anoplogasteridae, Anomalopi-
dae and Monocentridae are closely related and should be placed
in the suborder Trachichthyoidei, as suggested by Parr (1933).
The Gibberichthyidae, Stephanoberycidae and Melamphaidae
also appear closely related and form the suborder Slephanob-
erycoidei.
Despite the efforts of Rosen and Zehren, there are still prob-
lems with beryciform classification. Only adult characters have
presently been used, but early life history data is pertinent in
two instances. In the cladogram, a common ancestry is suggested
for the Diretmidae, Anoplogasteridae, Trachichthyidae. An-
omalopidae and Monocentridae with no character state to sep-
arate them. The larval head spine pattern in the Diretmidae and
Anoplogasteridae is similar and distinctive, and may help to
resolve the cladogram. Gibberichthys with its kasidoron stage
may appear to be vastly different from the Melamphaidae, but
the occurrence of very long branched pelvics in larval Poromitra
suggest a possible relationship (de Sylva and Eschmeyer, 1977).
In summary, further phylogenetic studies of the order Bery-
ciformes are needed in order to determine if the order is mono-
392
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 210. Kasidoron larva of Gibberichthys pumilus. 6.2 mm NL, DANA Sta. 3543 (source: de Sylva. pers. comm.).
Adioryx diodema
Holocentrus rufus
Plectrypops retrospinis
Ostichthys trachypoma
Myripristis sp.
Centroberyx affinis
4 — Beryx splendens
Scope log ad us mi zolepis
Scope logodus unispinis
Melamphaes macrocepholus
Melamphoes eulepis
Scopeloberyx sp.
Poromitra pumilus
Stephanoberyx monae
Gibberichthys pumilus
Diretmus orgenteus
Anopiogaster cornuto
Hoploslethus mediterroneus
Gephy robery X da r wi ni
Trachichthys australis
Paratrachichthys sp.
Photoblephoron palpebrotus
'Anomalops katopteron
Monocentrus japonicus
Cleidopus gtorioe- maris
Fig. 211. Cladogram showing the relationships of the beryciform
families studies by Zehren (1979).
phyletic, to determine the relationships between the various
suborders, and to determine the relationships of the Beryci-
formes to other orders of fishes. Inclusion of early life history
characters in these studies would be useful. However, the lack
of early Ufe history data for ten of the beryciform families may
prove a stumbling block in these efforts.
Division of Fishes, National Museum of Natural History,
Smithsonian Institution, Washington, District of
Columbia 20560.
Zeiformes: Development and Relationships
K. A. TiGHE AND M. J. KJEENE
THE order Zeiformes is diagnosed by a series of derived
characters that are not unique to the order (Heemstra.
1980): presence of dorsal spines in most forms; presence of anal
and pelvic spines in most forms; reduced number of pelvic and
caudal rays; absence of orbitosphenoid; absence of subocular
shelf; gills 3'/2 (no slit behind last hemibranch); mouth more or
less protrusible; no supramaxilla. Other characteristics of the
order as presented by Heemstra ( 1 980) are primitive characters
that shed little information on the relationships of the order.
The literature on Zeiformes is scattered and inadequate. Only
the family Zeidae has been examined on a world-wide basis
(Bray, 1983). Information on most species is descriptive, with
little known about ranges, life history stages, abundance, ecol-
ogy, and relationships. Zeiformes are marine and various species
occur in the tropical and temperate parts of all oceans in coastal,
benthic, epipelagic, mesopelagic, bathypelagic, and bathyben-
thic waters (Wheeler et al., 1973). Families are distinguished by
presence of vertically elongate or small or no scales, relative
body depth, relative mouth size, degree of development of anal
and pelvic fin spine(s), number of lateral lines, and morphology
of the eye-jaw region. Generic and specific designations are based
mainly on morphometnc, meristic, specialized scale, and color
characters (Heemstra, 1980).
The order Zeiformes is presently placed in superorder Acan-
thopterygii, near the Beryciformes and other groups that have
not attained the perciform level of structural organization.
Greenwood et al. (1966) included the Parazenidae, Grammi-
colepididae, Zeniontidae, Oreosomatidae, Zeidae, Caproidae,
and Macrurocyttidae in the Zeiformes. Heemstra (1980) revised
the Zeidae of South Africa and gives a key to all the zeiform
families above except the Caproidae which he, like earlier work-
ers (Rosen, 1973), feels is only superficially similar to zeiforms
and therefore should not be included in the order. He also pro-
vides diagnoses for the order and four of the remaining families.
Parazen pacificus, not reported from South Africa, is described
by Mead (1957). Keys to South African zeids and grammico-
lepidids are given by Heemstra ( 1 980), along with a key to adult
oreosomatids of the southern Atlantic and Indian Oceans sup-
plied to him by Karrer and Eschmeyer. Meristic ranges, number
of species, and number of genera for the six families presently
in the Zeiformes are given in Table 104.
Development
Early life history information on most zeiform species is non-
existent (Table 105). There is some information on prejuvenile
stages (specialized ontogenetic stages between larvae and ju-
veniles) for Oreosomatidae and Grammicolepididae, but none
on earlier stages. Early life history data for Zeus faber from egg
through juvenile is quite extensive, but such information is
incomplete or nonexistent for other zeid species. For the Cap-
roidae. larvae of Aniigonia capros and .-1. ruhescens are known,
as are all the early stages of Capros aper. Nothing is known for
the Parazenidae and Zeniontidae.
Eggs are known for two species of zeids. They are spherical.
have a single oil droplet, nonsegmented yolk, and a smooth
chorion. Eggs of Zeus faber range from 1 .8-2. 1 mm in diameter
with an oil droplet diameter of .32-.40 mm (Sanzo, 1956; De-
khnik, 1973; Robertson, 1975a). Those of Zenopsis nebulosus
are 2.0-2.25 mm with a droplet of .275-.375 mm (Robertson,
1975a). Eggs of Capros aper are about 1.0 mm in diameter,
spherical, and have a smooth chorion, unsegmented yolk and
a single oil droplet (Arbault and Boutin, 1968a; Sanzo, 1956).
Eggs of all other species of zeiform fishes are unknown.
Newly hatched larvae of Zeus faber were described by Sanzo
(1931b). Pigmentation is extensive over body, head and yolk
sac with the pigmentation extending to the margin of the dorsal
finfold and also on the base of the anal finfold for most of its
length (Fig. 212A). Only the tip of the caudal region is unpig-
mented. The pectoral and pelvic fin buds are present upon hatch-
ing. Preflexion larvae retain the extensive body pigmentation,
rapidly become deep-bodied, and show a precocious develop-
ment of the pelvic fins (Fig. 212B). Postflexion larvae have
almost all fin elements developed (Fig. 212C) and are rapidly
assuming the characters of the adult.
Larval stages are known for both genera in the family Cap-
roidae. Newly hatched larvae of Capros aper (Fig. 21 2D) have
large stellate melanophores on the dorsal, lateral and ventral
surface of the body with a few melanophores on the head and
associated with the oil globule. Preflexion larvae (Fig. 212E)
become very deep-bodied with an increase in head size. Pig-
mentation densely covers the entire body except for the caudal
region. A medial serrated ridge occurs on the cranium and other
paired serrate ridges develop along the lower jaw and in the
supraocular region. Numerous preopercular spines also develop
during this stage. Minute spines associated with the developing
scales cover the entire body (Fage, 1918). Transformation to
the juvenile is gradual and completed by a size of 15-20 mm
SL.
Larvae of Antigonia were described by Uchida (1936) and
Nakahara (1962). The larvae are relatively deep-bodied with
pigmentation on the peritoneum and head. The median serrate
cranial spine, serrate preopercular spines, and serrate ridges on
the frontal, mandibular and preopercular regions are character-
istic of both A. rubescens and A. capros (Fig. 213A, B), but are
totally lost before reaching juvenile sizes of 25 mm. There are
several differences between the larvae of the two species but the
most obvious is the presence of a vertically directed spine in
the occipital region of A. rubescens.
At least some grammicolepidids exhibit striking proportional
changes during growth. Smaller Grammicolepis brachtusculus
are very deep-bodied relative to larger ones based on an ex-
amination of specimens 70 to 400 mm SL (Quero, 1979). Young
Xenolepidichihys dalgleishi also have a relatively deeper body
than larger specimens (Myers, 1937) and possess long filamen-
tous extensions on some of the dorsal spines and on the first
anal spine (Smith, 1949; Fig. 279). These shorten greatly with
growth as shown by Myers' (1937) 71 mm SL specimen.
Oreosomatid adults have mainly overlapping cycloid or cte-
393
394
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 104. Meristic Ranges, Number of Genera, and Number of Species for the Families Placed in the Zeiformes by Greenwood et
AL. (1966). All data are from Heemstra (1980) unless noted.
Branched
caudal
Branchi-
Number of
Number of
Pelvic
Dorsal
Anal
rays
ostegals
Vertebrae
Pecloral
genera
species
Grammicolepididae
I, 6
V-VII, 27-34
II, 27-35
13
7
37-46
13-16«
3
5
Zeidae'
0-1, 6-10
VII-X, 22-37
I-IV, 20-39
11
7
29-42
11-14,
17-18
6
13
Oreosomatidae-
I, 5-6
V-VIII, 29-35
II-IV, 27-34
11
7
(3
genera)'
35-43
17-22
4(5?)
10
Zeniontidae
1.5-6
VI-VII, 25-29
O-II, 23-32
11
7-8
25-27
17"
3?
4?
Parazenidae'
0, 7
VIII, 26-30
I, 31-33
9
7
34
15-16
1
1
Caproidae"
I, 5
Vll-X, 26-37
III, 23-34
10
5-6
21-23'
I.
11-14
2
6
Macrurocyttidae'
I, 3
V, 27
T>
7
ca. 5
7
15
1
1
' Bray, 1983.
- Karrer and Eschmeyer, Ms.
'Mead. 1957.
' Berry. 1959a.
> Fowler. 1934.
•McAlhsIer. 1968.
' Rosen, 1973.
" Myers, 1937.
'Quero, 1978.
noid scales while the pelagic prejuveniles are oval in outline and
possess a leathery skin with distinct hardened cones or scaley
knobs laterally and ventrally (Myers, 1960; Eschmeyer et al.,
1 983). This stage is exhibited by Abe and Kaji's ( 1 972) specimen
of Oreosoma atlanlicum (Fig. 2 1 3C). Karrer and Eschmeyer (in
press) report prejuveniles of Pseudocytlus as large as 100 mm
and suggest that metamorphosis can be delayed. In one species,
the transformation is incomplete and the species remains in the
midwater prejuvenile habitat and becomes mature there.
In the Zeidae. only Cyttus traversi is presently known to have
a prejuvenile stage. This stage has a relatively much deeper body
than the adult, and bears long filamentous extensions with nu-
merous appendages from the dorsal spines and pelvic spine and
rays (James, 1976b: fig. 1). The prejuvenile stage occurs near
the surface in coastal waters (James, 1976b) while specimens
greater than about 100 mm have been caught near the bottom
at depths from 200 to 800 meters (Heemstra, 1980).
Macrurocytlus acanthopodiis was described by Fowler ( 1 934)
and was placed in the order Zeiformes by Greenwood et al.
(1966). Its small size, elongate pelvic spine, and stout dorsal
spine (Fig. 21 3D) array suggest that this may be a juvenile or
prejuvenile form, perhaps of the Zeniontidae as suggested by
Heemstra (1980).
Relationships
The present classification of the Zeiformes is based only on
characters of the adults. Heemstra (1980) includes five families
in the Zeiformes but speculates that the Grammicolepididae
Table 105. References Giving Descriptions and/or Illustrations of Eggs, Larvae and Prejuveniles of the Order Zeiformes.
Family
Eggs
Pre-juveniles
Grammicolepididae
Zeidae
Caproidae
Oreosomatidae
Zeniontidae
Parazenidae
Macrurocyttidae
Sanzo, 1931b
Sanzo, 1956
Dekhnik, 1973
Robertson, 1975a
Cunningham, 1889
Holt, 1897, 1899
Sanzo, 1956
Arbault and Boutin, 1968a
Ehrenbaum, 1905-1909
Schmidt, 1908
Sanzo, 1931b
Sanzo, 1956
Banarescu, 1964
Crossland, 1982
Holt, 1897, 1899
HefTord, 1910
Clark, 1914
Fage, 1918
Sanzo, 1956
Uchida, 1936
Nakahara, 1962
Smith, 1949
James, 1976b
Cuvier, 1829
Abe, 1957
Kobayashi et al., 1968
Abe and Kaji. 1972
Fowler, 1934
TIGHE AND KEENE: ZEIFORMES
395
Fig. 212. Zeiform larvae. (A) Yolk-sac larva of Zeus faber, 4.3 mm NL (source: Sanzo, 1931b); (B) Preflexion larva of Zeus faber. 4.3 mm
NL (source: Crossland, 1982); (C) Postflexion larva of Zeusfaher, 7.2 mm SL (source: Crossland. 1982); (D) Yolk-sac larva of Capros aper. 2.9
mm NL (source: Sanzo. 1956); and (E) Preflexion larva of Capros aper. 5.0 mm NL (source: Sanzo, 1956).
396
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 213. Zeiform larvae. (A) Preflexion larva of Anligonia rubescens. 4.5 mm TL (source: Uchida, 1936); (B) Poslflexion larva of Anligonia
capros, 4.75 mm TL (source; Nakahara, 1962); (C) Prejuvenile of Oreosoma allanlicum. 61 mm SL (source: Abe and K.aji, 1972); and (D)
Holotype o( Macrurocyttus acanthopodus Fowler 1934, 43 mm SL (source: Fowler, 1934).
TIGHE AND KEENE: ZEIFORMES
397
may prove to be incorrectly placed there because they differ
considerably in the configuration of their jaw elements, scales,
number of vertabrae. and have a higher number of caudal rays.
Heemstra's decision to exclude the Caproidae from the Zei-
formes is supported by evidence from Rosen (1973), who dis-
cusses some similarity between zeoids and caproids but states
that the pelvic count of 1 spine and 5 rays, 3 anal spines, and
the reduced vertebral number 21-23 are a combination of char-
398
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
acters found among percoids. The possession of normal abdom-
inal parapophyses, lack of ventral ridge scales or bucklers, and
a percoid type of caudal skeleton suggest to Rosen that caproids
appear to fit the present definition of a perciform while other
zeoids do not.
These findings support the movement of the Caproidae higher
in Acanthopterygian classification. The very different larvae of
the two caproid genera suggest that a thorough reexamination
of the osteology of adult representatives of these genera could
be necessary before the family is placed somewhere else.
There has been no phylogenetic systematic study of the order
Zeiformes. Inclusion of early life history characters would prob-
ably be useful in such a study, but these are unknown for most
members of the order.
group. Rosen has suggested that the Zeiformes do not represent
a monophyletic lineage, but are best included within the Te-
traodontiformes with which they are united by seven synapo-
morphies. Within Rosen's classification, the Caproidae are the
sister group to the rest of the Tetraodontiformes. In addition,
the rest of the zeiform families are united with the plectognath
fishes by four synapomorphies while the plectognath families
are monophyletic on the basis of six synapomorphies. Evidence
from early life history characters supporting this classification
is very limited due to the lack of knowledge of the early life
history of most of these fishes, but the similarity in morphology
and pigmentation between newly hatched Zens faher larvae and
tetraodontid larvae does provide some support for Rosen's hy-
pothesis.
Addendum: After this paper went to press. Rosen (1984) pub-
lished a phylogenetic analysis of the families (except Macru-
rocyttidae) herein included in the order Zeiformes which re-
sulted in a drastic change in the systematic placement of this
Division of Fishes, National Museum of Natural History,
Smithsonian Institution, Washington, District of
Columbia 20560.
Gasterosteiformes: Development and Relationships
R. A. Fritzsche
THE actinopterygian fish order Gasterosteiformes contains a
diverse assemblage of specialized fishes. There are about
220 species arranged into 10 or II families (Fritzsche. 1982).
Historically this group has been divided into two or three orders,
under such names as Lophobranchii, Thoracostei, Solenich-
thyes, Catosteomi, Hemibranchii, Hypostomides, Gasterostei-
formes, Syngnathiformes, and Pegasiformes (Boulenger, 1904;
Berg, 1940; and Starks, 1902). Pietsch (1978b) presented infor-
mation which suggests that Pegasiformes are intermediate be-
tween the Gasterosteiformes and Syngnathiformes. Pegasids are
intermediate in (1) snout development and in the condition of
the nasal bones; (2) retention of the parietals; (3) retention of
three circumorbital bones; (4) presence of a dorsal strut join-
ing the ceratohyal and epihyal; (5) reduction in number of ele-
ments of the branchial arches; (6) the presence of two pairs of
pleural ribs; and (7) retention of support for a spinous dorsal
fin (Pietsch, 1978b). He proposed a tentative classification unit-
ing all three groups into the single order Gasterosteiformes. This
order is characterized by (1) branchiostegal rays reduced to 1-
5; (2) absence of supramaxillary, orbitosphenoid, and basi-
sphenoid; (3) postcleithrum reduced to single bone or absent;
(4) pelvic girdle never attached directly to cleithra; (5) rather
small mouth, often at end of more or less tubular snout; and
(6) armor of dermal plates covers most members (Fritzsche,
1982). Pegasids form the primitive sister-group of the Soleno-
stomidae and Syngnathidae. These families share a number of
derived character states including (1) feeding mechanism; (2)
metapterygoid absent; (3) hyoid apparatus short, bearing elon-
gate, filamentous branchiostegal rays; (4) gill opening restricted
to a small hole on the dorsolateral surface behind head; (5) gill
filaments tufted or lobe-like; (6) articular processes of mobile
vertebral centra absent; (7) posttemporal co-ossified with cra-
nium; (8) postcleithrum absent; and (9) head and trunk encased
by bony plates, tail encircled by bony rings (Pietsch, 1978b).
The Pegasidae, Solenostomidae and Syngnathidae form the
primitive sister-group of the Macrorhamphosidae, Centriscidae,
Aulostomidae, and Fistulariidae and the resulting classification
is as follows:
Order Gasterosteiformes
Suborder Gasterosteoidei
Superfamily Aulorhynchoidea
Family Aulorhynchidae
B
Fig. 214. Eggs of some gasterosteiforms; (A) Gasterosleus acuteatus
(from Kuntz and RadclifTe, 1917); (B) Fistularta pelimha (from Mito,
1 96 1 a); (C) Macrorhamphosus scolopa.x (horn Hardy, 1 978a, after Spar-
ta, 1936); (D) l/ippocumpus ereclus (from Hardy, 1978a).
FRITZSCHE: GASTEROSTEIFORMES
399
B
"T-^Mcji.^ -,■<:,; -.-v.v? >»,'.c v>7.:-
j*tf»i^ iJo^T-vsrrrr
-■■*g
Fig. 215. Larvae of some gasterosteoids. (A, B) Aulorhynchus flavidus. 8 mm TL and 23 mm TL (from Marliave, 1 976); (C) Apeltes quadracus.
6.0 mm TL (from Ryder, 1887); (D) Apeltes quadracus. 10.5 mm TL (from Hardy, 1978a).
Family Hypoptychidae
Superfamily Gasterosteoidea
Family Gasterostcidae
Suborder Syngnathoidei
Infraorder Syngnatha
Superfamily Pegasoidea
Family Pegasidae
Superfamily Syngnathoidea
Family Solenostomidae
Family Syngnathidae
Infraorder Macrorhamphosa
Superfamily Macrorhamphosoidea
Family Macrorhamphosidae
Family Centriscidae
Superfamily Aulostomoidea
Family Aulostomidae
Family Fistulariidae
The taxonomy within this order is poorly understood. The
lack of agreement regarding relationships within the Gasteros-
teiis aculeatus complex (Bell, 1976) and whether or not Macro-
rhamphosus contains only one species (Ehrich, 1 976) are two
examples. Recent studies, such as that of Fritzsche (1980), have
shown that many species of syngnalhids are morphologically
plastic. This plasticity has been the cause of a proliferation of
species and subspecies descriptions in the literature. The process
of sorting out the nominal species still continues for most taxa
included in Gasterosteiformes.
Gasterosteiforms are found in freshwater, estuarine, and ma-
rine habitats through tropical and temperate regions. Most species
are relatively small and cryptically colored. They have no real
fishery importance and usually are thought of as interesting
aquarium fishes or simply curiosities, e.g. the seahorse. Since
commercial importance is lacking, there is very little literature
dealing with the early life histories of these fishes except for
400
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
. ,^ i_i — i — L-iLJLJ)
Fig. 216. Larvae of some pegasoids and syngnathoids. (A) Pegasidae, 2.4 mm (from Leis and Rennis, 1983); (B) Solenoslomus sp., 5.1 mm
NL (ongmal illustration by Wayne A. Laroche); (C) Syngnathus fuscus. ca. 3.5 mm TL (from Ryder, 1887); (D) Hippocampus japomcus. ca. 6
mm TL (from Nakamura, 1937).
anecdotal accounts or descriptions of chance collections of eggs
or young.
Development
There are published descriptions of the eggs of Aulorhynchus
(Limbaugh, 1962; Ida, 1976), Hypoptychiis(\s\\i%aku 1957; Ida,
1976), gasterosteids (notably Kuntz and Radclifle, 1917; Vrat,
1949; Swamp, 1958), Solenoslomus (Padmanabhan, 1961),
Macrorhamphosus (Sparta, 1936), and Fistulana (Delsman,
1921; Mito, 1961a; Watson and Leis, 1974). There are few
descriptions of the eggs of syngnathids due to the unique male
brooding habits of this group, however, Hudson and Hardy
(1975) provided a good description of Hippocampus erectus
eggs. Most accounts simply include the number and size of eggs
in the male's pouch (e.g., Fritzsche, 1980). Gudger (1905) pro-
vided a fairly extensive treatment of the embryology of Syng-
nathus floridae.
Larvae (usually just one or two and not a series) have been
described, for .-l«/o/-/;i«c/!i« (Limbaugh, 1962; Marliave, 1976),
gasterosteids (Kuntz and Radcliffe, 1917; Vrat, 1949; Swarup,
1958), pegasids (Jones and Pantulu, 1 958; Jones and Kumaran,
1967; Leis and Rennis, 1983), Solenoslomus (Padmanabhan,
FRITZSCHE: GASTEROSTEIFORMES
401
B
Fig. 217. Larvae of some macrorhamphosoids and some aulostomoids. (A) Macrorhamphosus scolopax. 3.0 mm TL (from Hardy, 1978a,
after Sparta, 1936); (B) Centriscidae. 2.7 mm (from Leis and Rennis. 1983); (C) Fistularia petimba. 7.08 mm (from Mito, 1961a).
1961), syngnathids (most notably D'Ancona, 1 933c; Nakamura,
1937; Takai and Mizokami, 1959; James, 1970; Russell, 1976;
Dawson et al., 1979), macrorhamphosids (D'Ancona, 1933d;
Sparta, 1936; Mohr, 1937), centriscids (Mohr, 1937; Leis and
Rennis, 1983) and F/5n//ana (Jungersen, 1910; Delsman, 1921;
Mito, 1961a; Leis and Rennis, 1983). Larvae have not been
described for Hypoplychus and Aulostoinus.
Osteological development has not been studied for most gas-
Table 106. Meristic Characters for Families of the Gasterosteiformes (adapted from Pietsch, 1978b).
Hypo.
ptychi-
Oaster-
Soleno-
Syng-
Macrorham-
Fislu-
Character
Aulorhynchidae
dae
ostcidac
Pegasidac
stomidac
nathidae
phosidae
Cenlnscidae
Aulostomidae
lanidae
Circumorbital bones
3
T
3
3
0
2-3
1
1 ( + 2'')
1 (O'^)
0
Branchiostegal rays
4
4
3-4
5
1 (bifid)
1-3
4-5
3-5
4
5
Vertebrae
52-56
55
28-42
19-22
33
37-77 -h
24
20
59-64
76-87
Elongate anterior
0
0
0
6
3
3
5
5-6
4
4
vertebrae
Pleural nbs (pairs)
0-22
9
9-16
-)
0
0
0
0
0
0
Dorsal-fin rays
XXIV-XXVI
+ 9-10
20
III-XVI
-1- 6-14
5
V -h 18-23
0-60
IV-VII
+ 9-11
III
+ 10-12
VIII-XIII
-1- 21-26
14-20
Anal-fin rays
I -1- 9-10
20
I -F 6-12
5
16-23
0-6
19-20
11-12
22-27
14-19
Caudal-fin rays
13
13
11-12
8
16
0-11
23
11
20
22-24
Pectoral-fin rays
10-11
9
9-23
10-18
18-27
0-23
15
10-12
15-16
13-18
Pelvic-fin rays
1 + 4
0
I + 0-2
I + 2-3
I +6
0
1 + 4
I + 4
6
5-6
402
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 218. Larvae of some pegasoids and syngnathoids. (A) Eurypegasus papilio. 7.0 mm (from Leis and Rennis, 1983); (B) Solenoslomus sp.,
11.5 mm SL (original illustration by Wayne A. Laroche); (C) Syngnathus schlegeli. size unknown (from Chyung, 1977); (D) Yozia bicoarctata.
ca. 10-1 1 mm SL (from Dawson et al., 1979); (E) Hippocampus japomcus. ca. 6.5 mm TL (from Nakamura, 1937).
FRITZSCHE: GASTEROSTEIFORMES
403
B
Fig. 219. Larvae of some macrorhamphosoids and aulostomoids. (A) Macrorhamphosus scolopax. 9.0 mm TL (from Hardy, 1978a, after
D'Ancona, \97>'!i(X)\(Q) Aeoliscus stngalus. 7.9 mm (from Leisand Rennis, \9S3y, {C) Fisliilaria pelimba. 15.6 mm (from Leis and Rennis, 1983).
terosteiforms. Kindred (1921) presented a classic study on the
chondrocranium of Syngnalhus fusciis. Padmanabhan (1961)
published information on the development of jaws in Solenos-
tomus cyanopterus. Development of the bony rings on the body
ofSyngnaihus typhle was studied in detail by Czolowska ( 1 962).
Considering the diversity of habitats and spawning behaviors
found within the group, it is difficult to identify a character or
suite of characters that typifies all members of this order. Some
gasterosteiforms spawn in open water and produce buoyant eggs
(e.g., Fistularia. Watson and Leis, 1974); others such as the
sticklebacks and tubesnouts (Gasterosteidae and Aulorhynchi-
dae) construct nests out of vegetation for receipt of the eggs;
while others such as the seahorses, pipefishes, and ghost pipe-
fishes (Syngnathidae and Solenostomidae) brood the eggs within
specialized structures located on one of the parents. Syngnathids
have a most unusual adaptation in having a specialized patch
or pouch (marsupium) developed on the males for receipt and
incubation of eggs. Those groups containing species that broad-
cast spawn or have nests produce larvae that go through the
typical developmental pattern of pelagic larvae. Those that brood
eggs, such as the more advanced syngnathids, may retain the
eggs and developing larvae until the young have reached a ju-
venile stage of development.
In general, eggs of most gasterosteiforms are spherical, how-
ever, those of Hippocampus have been described as being dis-
tinctly pear-shaped (Hudson and Hardy. 1975) or ellipsoidal
(Nakamura, 1937) (Fig. 214). The eggs typically have numerous
oil droplets in the yolk (Gudger, 1905; Kuntz and Radcliffe,
1917). However, those of Fistularia lack oil droplets (Watson
and Leis, 1974), and Macrorhamphosus has a single oil globule
(Lo Bianco, 1909; Fage, 1918). The perivitelline space is narrow
in Solenostomus (Padmanabhan, 1961), gasterosteids (Hardy,
1978), and Fistularia (Mito, 1961a), while it is relatively wide
in Hippocampus (Hardy, 1 978a). The yolk is not segmented and
is typically yellow in syngnathids (James, 1970), rose- violet in
Macrorhamphosus (Hardy, 1 978a), and clear in Fistularia (Mito,
1961a). The chorion is typically smooth, however, small at-
tachment threads have been reported for some gasterosteids
(Hardy, 1978a). Most gasterosteiforms have eggs about 1.0mm
in diameter except that Solenostomus eggs are about 0.6 mm
(Padmanabhan, 1961) and Hippocampus eggs may approach
4.0 mm in one dimension (Hardy, 1978a).
Larvae of most gasterosteiforms (except gasterosteids) have
a very distinctive, elongate snout bearing a small upturned mouth
which reflects a trenchant character of the adults (Figs. 2 1 5-
219). Meristic characters are quite variable in this order (Table
106). Myomere counts range from a low of 19 in pegasids to 87
in Fistularia (Leis and Rennis, 1983). Fin ray meristics are
equally variable and some groups lack one or all of the fins
(Table 106). Syngnathids, for example, may have 0 to 60 dorsal
404
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
fin rays. Size at hatching has not been well documented for
gasterosteiforms. Gasterosteids and Solenostomus may hatch at
3.0 mm TL (Padmanabhan, 1961; Hardy, 1978a), while Aulo-
rhynchus hatch at 5.5-8.0 mm TL (Marliave, 1976). Presence
of bony plates rather than scales is the rule in this order. These
plates are typically present and easily seen by the time notochord
flexion is complete (Figs. 218 and 219). Several groups develop
small spinules in the skin on the body early in development.
Macrorhamphosus develops spinules at about 6 mm TL (Sparta,
1936). All species of Fisliilaha go through a so-called "villosa
stage" (Liitken, 1880) during which they are covered with small
spinules (Fig. 219C). Pigmentation of species for which larvae
have been described varies from very heavy pigmentation in
Gasterosteidae and Macrorhamphosidae to rather light pig-
mentation in Syngnathidae and Fistulariidae. The young of sev-
eral species of syngnathids have conspicuous dark bars (D'An-
cona, 1933c; Takai and Mizokami, 1959; and Fritzsche, 1980)
(Fig. 2 1 8C). Dawson et al. ( 1979) reported the presence of elon-
gate dermal appendages in young of the syngnathid genus Yozia
(Fig. 218D). They believed that these appendages have a buoy-
ant function for aid in distribution of the pelagic young.
Relationships
Besides the hypothesis of relationships proposed by Pietsch
(1978b), there are several other recent hypotheses. Greenwood,
et al. (1966) proposed the following classification scheme:
Order Gasterosteiformes
Suborder Gasterosteoidei
Family Gasterosteidae
Family Aulorhynchidae
Family Indostomidae
Suborder Aulostomoidei
Family Aulostomidae
Family Fistulariidae
Family Macrorhamphosidae
Family Centriscidae
Suborder Syngnathoidei
Family Solenostomidae
Family Syngnathidae
Order Pegasiformes
Family Pegasidae
The family Indostomidae has at various times been thought to
be related to the gasterosteiforms (Bolin, 1936b; Berg, 1940).
But, Pietsch (1978b) has pointed out that the specific relation-
ship of this family must await further investigation. I have,
therefore, not included this monotypic family (I ndostoiniis par-
adoxus) in this account.
Banister (1967) proposed a classification based on his osteo-
logical studies as follows:
Order Aulorhynchiformes
Family Aulorhynchidae
Family Gasterosteidae
Order Aulostomiformes
Suborder Aulostomoidei
Family Aulostomidae
Family Fistulariidae
Family Solenostomidae
Family Syngnathidae
Suborder Centriscoidei
Family Macrorhamphosidae
Family Centriscidae
His scheme differs little from previous ideas except in use of
new ordinal names (to reduce confusion?) and inclusion of the
closely related macrorhamphosids and centriscids in their own
suborder. Characters of his Centriscoidei are ( I ) separate meta-
pterygoid present and anterior end of quadrate normal; (2) nasals
large and elongated; (3) five or more modified anterior vertebrae;
(4) supraethmoid contributes little to dorsum of snout; (5) post-
temporal pyramidal; (6) caudal fin skeleton uniform, with single
large hypural plate; (7) vertebral number low (about 20); (8) no
sign of reduction in pharyngeal skeleton; and (9) intemeurals
for vertebrae five and six absent. Banister's (1967) hypothesis
of relationships has not been published.
Nelson (1976) proposed a classification that was similar to
that of Greenwood et al. (1966) except that the families Gas-
terosteidae and Aulorhynchidae were recognized as forming the
order Gasterosteiformes while the remainder of the families
were placed in Syngnathiformes. This separation was done
pending clarification of relationships and establishment of
monophyly. As noted earlier, Pietsch (1978b) was able to link
the two groups based on the intermediate nature of the pega-
soids.
Ida (1976) demonstrated that the monotypic Hypopiychus
dybowskii Steindachner resembled gasterosteids and aulorhyn-
chids in osteology, mode of life, and reproduction. He, therefore,
removed this species from the Perciformes and placed it close
to the Gasterosteidae and Aulorhynchidae in the suborder Gas-
terosteoidei of his order Syngnathiformes.
Early life history stages have contributed little to the devel-
opment of the above hypotheses of relationships. Pietsch ( 1 978b)
showed that snout structure of Pegasus and Macrorhamphosus
is very much alike at small sizes even though it is quite different
in adults. Ida (1976) used egg morphology as one of the char-
acters supporting his placement of the Hypoptychidae close to
the Gasterosteidae.
Considering the paucity of developmental descriptions for
species of the Gasterosteiformes, it is difficult to test existing
hypotheses of relationships using developmental characters.
However, it is interesting to note the sequence of fin formation
seems to support the close relationship of the Gasterosteidae
and Aulorhynchidae. Aulorhynchus forms the pectoral fins first,
followed by the caudal, second dorsal and anal fins (Marliave,
1976). The gasterosteid Apeltes follows the same sequence
(Hardy, 1978a). Gasterosteus forms the pectoral fins after the
anal fin (Hardy, 1978a). Few developmental sequences are known
for the other gasterosteiforms. Those that are available show
that for the pegasids, macrorhamphosids and syngnathids the
sequence begins with the development of the dorsal fin followed
by the anal, caudal and pectoral. It may well be that the sequence
of fin formation will provide evidence for the retention of the
Gasterosteidae and Aulorhynchidae in their own order or sub-
order. Additionally Macrorhamphosus. Acoliscus and Fislularia
develop a dorsal finfold that extends on to the head which might
be given as evidence in support of Pietsch's (1978b) infraorder
Macrorhamphosa. However, pegasids also have this anteriorly
placed finfold (Leis and Rennis, 1983). This coupled with the
low myomere numbers for pegasids and macrorhamphosids may
indicate that these two groups should be placed closer together
than is presently indicated in Pietsch's treatment. This question
must remain unresolved pending further descriptive and com-
parative work on gasterosteiform larvae.
Studies of the relationships of Gasterosteiformes to other taxa
have been dominated by unsupported hypotheses. Gosline(197I)
proposed that the "origin for both gasterosteoids and synga-
FRITZSCHE: GASTEROSTEIFORMES
405
thoids (sic) suggest one or two origins in the percopsiform —
beryciform area." The mixture of advanced and primitive char-
acters shown by gasterosteiforms suggested to Banister (1967)
evolution "from a primitive myctophoid type offish . . . towards
an acanthopterygian grade." McAllister (1968) suggested "the
Gasterosteiformes are derivable from the Perciformes" and ". . .
the Syngnathiformes from the subperciforms, such as Beryci-
formes and Zeiformes." In fact he suggests that Antigonia or
Capromimus would appear to be close to the ancestors of the
Syngnathiformes. None of these authors presented evidence for
support of their ideas. Examination of the description o^ Anti-
gonia larvae by Nakahara ( 1 962) shows that this fish bears little
resemblance to the early stages of described gasterosteiforms.
Larval Antigonia are characterized by well-developed, serrated
preopercular and cranial spines. These spines are never seen in
gasterosteiform larvae. However, the description of the larvae
of Capros aper (Russell, 1976) indicates that the most charac-
teristic feature of them is the occurrence of small spines all over
the body surface. Additionally the larvae of C. aper are darkly
pigmented. These two characteristics are also found in some
gasterosteiform larvae, e.g., Macrorhamphosus. It is therefore
tempting to use these characters in support of McAllister's hy-
pothesis, however we will have to wait for further information
on both gasterosteiforms and zeiforms before we can support
or refute this hypothesis.
Department of Fisheries, Humboldt State University, Ar-
CATA, California 95521.
Scorpaeniformes: Development
B. B. Washington, H. G. Moser, W. A. Laroche and W. J. Richards
THE Scorpaeniformes are the fourth largest order of fishes
encompassing about 20 families (depending on classifi-
cation used), 250 genera and over 1 ,000 species. Representatives
of the order are widely distributed from tropical to arctic and
antarctic waters. Most scorpaeniforms are benthic or epibenthic
with representatives ranging from freshwater to the deep ocean.
The morphologically diverse "mail-cheeked fishes" are named
for the bony suborbital stay which extends posteriorly from the
third infraorbital to the preopercle. The suborbital stay is the
only known character that defines the order; however, some
workers have suggested that the stay evolved independently in
several lineages and may not indicate monophyly (Matsubara,
1943; Quast, 1965; Poss, 1975). The classification of the scor-
paeniforms is controversial, not only in terms of monophyly
but also at the subordinal and familial levels. Discussion of the
taxonomic status and current hypotheses of relationships is pre-
sented in Scorpaeniformes: Relationships (this volume).
Modes of reproduction vary widely within the scorpaeni-
forms. Many families spawn individual pelagic eggs (Anoplo-
pomatidae, Congiopodidae, Hoplichthyidae and Triglidae), while
others spawn demersal clusters of adhesive eggs ( Agonidae. Cot-
tidae, Cyclopteridae and Hexagrammidae). Where known, most
scorpaenids produce pelagic egg masses enclosed in a gelatinous
matrix. Notable exceptions include the scorpaenid genus Se-
bastes and the comephorids of Lake Baikal which give birth to
live young.
Larvae of only about 20% of scorpaeniform genera and ap-
proximately 10% of the species are known. Because of the wide
diversity of form, we are not able to characterize a typical scor-
paeniform larva. Early life stages of many scorpaeniforms are
characterized by strong head spination as depicted in the gen-
eralized scorpaenid larva Sebastes (Fig. 220). However, the
expression of head spination is variable within the order with
elaborations and losses in many groups.
For the purposes of this paper, we consider the Scorpaeni-
formes to be monophyletic and utilize the broad suborders Scor-
paenoidei and Cottoidei as a framework for presentation and
discussion. Because of the order's morphological diversity and
the lack of an agreed upon classification, discussion of larval
taxonomy is focused upon each family. The scorpaeniform fam-
ily Cyclopteridae is presented in the subsequent article in this
volume.
SCORPAENOIDEI
Eggs
Eggs are known for seven of the scorpaenoid families recog-
nized in Washington et al. (this volume), however, they are
known only for a few species (Table 107). Most scorpaenoid
families are oviparous and spawn pelagic eggs; however, repro-
ductive modes are varied in the Scorpaenidae. In the scorpaenid
subfamilies Scorpaeninae, Pteroinae, and Sebastolobinae the
eggs are extruded in bilobed gelatinous egg masses which float
at the surface. The eggs are slightly elliptical and have homo-
geneous yolk, a narrow perivitelline space, and a smooth cho-
rion. A single oil globule is present in Pterois (0.16-0.17 mm)
and Sebastolobus (0. 18-0.20 mm); Scorpaena lacks an oil glob-
ule. In the choridactyline genus Inimicus, eggs are extruded
singly, are spherical, and lack an oil globule (Table 107). Mem-
bers of the scorpaenid subfamily Sebastinae are viviparous and
give birth to large broods of young which are comparable in
stage of development to first-feeding larvae of oviparous scor-
paenids. The eggs are retained in the lumen of the ovary after
ovulation, range between 0.75 and 1.9 mm, have homogeneous
yolk, a narrow perivitelline space, smooth chorion, and one to
many oil globules. For the other families for which eggs are
known, the eggs are pelagic with none to multiple oil globules
(Table 107).
Larvae
At least one larval stage is known for 64 of the more than
600 species of scorpaenoids and for 20 of the 100+ genera.
Major reviews of larval scorpaenoids include Sparta (1956b)
406
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 107. Summary of Eggs and Larval Size Characteristics of the Scorpaeniformes based on Available Literature (excluding
Cyclopteridae).
Family/subfamily
species
Type of egg
pelagic (P).
demersal
(D) or vivip-
arous (V)
Egg size
(mm)
Number
of oil
globules
Largest oil
globule size
(mm)
Body length (mm) at
Hatchmg
Transfor-
mation
SCORPAENOIDEI
Scorpaenidae
Sebastinae'
Sebastes capensis
S. fasciatus
V
V
-
-
S. marinus
V
1.5
-
S. viviparous
S. hubbsi
S. inermis
S. longispinis
V
V
V
V
1.36
1
S. marmoratus
S. nigricans
V
V
0.75-0.95
1.6
1
many to
1
S. oblongus
S. pachycephalus
V
V
1.56-1.60
1.5-1.9
many
many
S. schlegeU
S. steindachneri
S. taczanowskii
S. const ellal us
V
V
V
V
-
-
S. cortezi
S. crameri
V
V
—
—
S. dallii
S. entomelas
V
V
-
-
S. flavidus
S. helvomaculatus
S. jordani
S. levis
S. macdonaldi
S. melanops
V
V
V
V
S. melanostomus
V
S. ovalis
V
S. paucispinis
S. pinniger
V
V
S. rufus
V
S. zacentrus
V
Sebastes Type A
Helicolenis dactyloplerus
V
V
Scorpaeninae
Pontinus Type A
Pontinus Type B
0.2
1.0
ca. 0.20
3.8
ca. 5.8
6.7-7.2
5.4-5.8
ca. 4.4
4.5
5.8-6.1
ca. 4,5
6,9-7.0
7.2-7.5
6.0-7.0
ca. 6. 1
ca, 4,8
ca, 5.4
4.0-5.0
4.1
ca. 5.7
5.0
4,5-
4.5
4.1
5,4
5.0
4.0-
4.6
5.0
4.5
4.9-5.1
4.6
4.0
4.6-4.8
ca. 4.3
4.2
2.2
<2.3
6.2-7.0
8.5-10.0
7.i
ca. 6
ca. 7
6.4-7
ca, 8
ca.
ca.
8.5
<7.1
7.0-8.3
8.0-9.3
6.2-8.0
9.9-12.9
7.7-8.0
6.2-7,2
ca. 6.8
7.2-9.7
ca. 7.8
6.1-7.6
7.4-8.5
7.0-
6.0-
■7.6
■7.9
ca. 20
8.5-11.8 ca. 24
10.6 -
ca. 18
ca. 17
ca. 10
12-14
>13
ca. 17
16-21
<20
21.7-30.6
23.6-26.7
12.0-18.6
8.0-10.0
27-30
7.6-10.4
ca. 19
7.7-9.0
ca. 15
—
23,2-30.6
ca. 16
15
12.8-
18.4
13.7-19.6
>19
4.1-4.6 ca. 15
< 5.0-5.5 ca. 10
Moseret al.. 1977
Moseret al.. 1977; Fa-
hay, 1983
Moseret al., 1977; TSn-
mg, 1961
T4nmg, 1961
Uchida et al., 1958
Harada, 1962
Takai and Fukunaga,
1971
Tsukahara, 1962
Fujita, 1957b, 1959
Fujita, 1958
Shiokawa and Tsukahara,
1961
Sasaki, 1974
Sasaki, 1974
Sasaki, 1974
Moser and Butler, in
press
Moseret al., 1977
Westrheim, 1975; Rich-
ardson and Laroche.
1979
Moser and Butler, 1981
Laroche and Richardson,
1981; Moser and But-
ler, in press
DeLacy et al., 1964; Lar-
oche and Richardson,
1980
Richardson and Laroche.
1979; Westrheim, 1975
Moseret al., 1977
Moseret al.. 1977
Moseret al., 1977
Laroche and Richardson,
1980
Moser and Ahlstrom,
1978
Moser and Butler, in
press
Moseret al., 1977
Waldron, 1968; Richard-
son and Laroche, 1979
Moser and Butler, in
press
Laroche and Richardson,
1981; Westrheim, 1975
Moseret al., 1977
Graham, 1939; Sparta,
1956b; Tuning, 1961;
Moser et al,, 1977; Fa-
hay, 1983
Moseret al., 1977
Moseret al., 1977
WASHINGTON ET AL.: SCORPAENIFORMES
407
Table 107. Continued.
Type of egg
pelagic (P).
demersal
(D) or vivip
arous (V)
Egg si/e
(mm)
Number
of oil
globules
Largest oil
globule size
(mm)
Body length (mm) at
Family/subfamily
species
Hatching
FIcMon
Transfor-
mation
References
Scorpaena guttata
P
1.22-
1.29 X
1.16-
1.19
0
1.9-2.0
4.5-5.7
>13
DavKJ, 1939; Onon,
1955d; Moseret al.,
1977
S. notata
P
0.76 X 0.88
0
—
<2.7
ca. 6.0
—
Sparta, 1956b
S. porcus
P
0.84 X 0.92
0
—
1.72
ca. 6.7
ca. 12
Sparta, 1956b
S. scrofa
P
0.68 X 0.88
0
—
<2.8
ca. 6.0
ca. 17
Sparta, 1956b
Scorpaena Type A
Scorpaenodes xyris
-
-
-
-
ca. 2.0
1.8
4.0-5.5
4.0-5,4
>12
11-14
Moseret al., 1977
Moseret al., 1977
Pteroinae
Pterois lunulata
P
0.81-0.83
1
0.16-0.17
1.52-1.58
-
-
Mitoand Uchida. 1958;
Mito, 1963
Dendrochtrus brachypte-
P
-
0
-
ca. 1.1
-
-
Fishelson, 1975
rus
Sebastolobinae
Sebastolobus alascanus
S. allivelis
Setarchinae
Eclreposehastes imus
Choridactylinae
Inimicus japonicus
Minoinae
Minous sp. (?)
Triglidae
Chelidonichthys cuculus
C. gurnardus
C. kumu
C. lastoviza
C. lucerna
C. obscurus
Lepidotrigla alata
L. aspera
L. japonica
L. microptera
Prionotus carolinus
P. evolans
Peristediidae
Peristedion
Congiopodidae
Congiopodus
leucopaecilus^
C. spinifer^
C. torvus^
Platycephalidae
Platycephalus indicus^
Platycephalidae spp.
P
P
1,2-1.4
1.2-1.4
1.31-1.43
0.18-0.20 ca. 2.6
0.18-0.20 ca. 2.6
<2.8
6.0-7.3
6.0-7.3
ca. 5.5
3.18-3.27 6.4-8.2
ca. 1.8
3.7-5,9
14-20 Pearcy. 1962; Moser,
1974; Moser et al.,
1977
14-20 Moser etal., 1977
ca. 28 Moseret al., 1977
ca. 10.4 Fujita and Nakahara,
1955; Mito, 1963; Sha
etal., 1981
>9.0 Leis and Rennis, 1983
p
1,45-1,65 1
p
1,45-1,5 1
p
1,20-1.27 1
p
1.29-1.33 1
p
1.25-1.36 1
p
—
p
1,22-1.25 1
p
1,16 1
p
1,20-1.40 1
p
1.26-1.31 1
p
0.94-1,15 1
0,19-0,33 3,2 9,0
0.25-0,27 3,12-3,26
0,24 - -
0,26-0,28 3.2 9.0
0.25 2.78-2,92
0.21-0.22 3.2 7.0
0.25-0.26 -
0.25-0.28 -
0-25 - 2.6-2.8
- 6.3
—
Padoa, 1956e
17.0
Padoa, 1956e
—
Uchida etal., 1958
—
Padoa, 1956e
17.0
Padoa. 1956e
—
Padoa, 1956e
—
Mho, 1963
19.0
Padoa, 1956e
—
Mito, 1963
—
Mito, 1963
8.6
Fritzsche, 1978; Fahay
1983
8.7
Fahay, 1983
1.7
1.9-2.2
1.82
<11.5
NA
NA
5-6
15.0 Padoa, 1956e; Breder and
Rosen, 1966
- Robertson, 1974, 1975a
>12.4 Brownell, 1979; Gilchrist,
1904; Gilchrist and
Hunter, 1919; Robert-
son, 1975a
p?
1.7-1.8
0
NA
—
—
—
Gilchrist, 1904
p
0.88-1.2
1
0.19-0.25
1.78-2.3
-2.1
7.3
3.9-5.2
13
Ueno and Fujita, 1958;
Change! al.. 1980
Uisand Rennis, 1983
408
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 107. Continued.
Family/subfamily
species
Type of egg
pelagic (P),
demersal
(D)orvivip- Egg size
arous (V) (mm)
Number Largest oil
of oil globule size
globules (mm)
Body length (mm) at
Hatching
Transfor-
mation
Hoplichthyidae
Hoplichthys haswellP
Hoplichlhys sp.^
Dactylopteridae
Daclylopterus volitans
Daicocus petersoni
Dactyloptena sp.
0.85-0.90 1
-0.8
0.15
0.14
e8
Robertson, 1975a
Okiyama (unpubl. MS)
<7
ca.
16
Fntzsche, 1978; Sanzo,
1933c; Padoa, 1956e
4.3
ca.
10
Senta, 1958
3.9-6.5
ca.
10
Leisand Rennis, 1983
COTTOIDEI
Agonidae
Agonomalus mozinoi
D
-1.0
—
—
5.5
Agonopsis chiloensis^
—
—
—
—
—
Agonus cataphractus
D
1.7-2.2
Several
co-
alesce
0.7-0.75
6.3-8.0
A. decagonus-'
—
—
—
—
—
Aspidophowides
—
—
—
—
—
monopterigius^
A. olriki'
—
—
—
—
—
Bolhragonus swani
D
2
—
—
7.5
Pallasina barbata-
—
—
—
—
—
Xeneretmus latifrons'
-
-
—
—
~7
Anoplopomatidae
A noplopoma fimbria
Blepsias cinhosus- —
Chitonotus pugetensis^ D
Clinocottus acuticeps D
C. analis^ D
C. embryum —
C. globiceps D
C. recalvus D
Cottus asper D
C bairdi D
C. carolinae D
C. cognatus D
C nozawae D
C. reinii^ D
2.0-2.1 -
Comephoridae
Comephoms baicalensis
C. dybowskii
V
V
N/A
N/A
Cottidae
Artedius creaseri
A. fenestralis
-
:
A. harringtoni
A. lateralis
D
1.07
A. meanyi
Ascelichthys rhodorus
D
1.7-
0.22
1.02-1.05 1 large 0.3
5-8 small
1.0-1.2 - -
1.2-1.3 several 0.18
large
1.5-2.0 - -
1.25-1.35 -
1-3
2.6-3.3
-10-12
-10
11-14
— Marliave, 1978
— de Ciechomski, 1981
— 14 mm Russell, 1976; Ehren-
baum, 1904; Mcintosh
and Pnnce, 1890
— Ehrenbaum, 1905-1909
— Dannevig, 1919; Bigelow
and Schroeder, 1953
— Dunbar, 1947
>16 Marliave, 1975
— Marliave, 1975
— Marliave, 1975
>33 Ahlstrom and Stevens,
1976; Hart, 1973; Ko-
bayashi, 1957
3.1-3.5
2.0-2.6
9.4
—
>48
Chemyayev, 1975
8.2
-13
-21
Chemyayev, 1971
-3.5
5.7-7.9
13-14
Washington, 1981
3.5-3.8
5.9-6.8
12-13
Washmgton. 1981; Mar-
liave, 1975
-3.0
5.2-6.4
12-14
Washmgton, 1981
3.9-4.5
5.0-6.3
9.5-10.5
Washmgton, 1981; Mar-
liave, 1975; Budd,
1940
-3
6.3-9.4
15-20
Washington, 1981
6.0
8.8-9.0
12-15
Malarese and Marliave.
1982
—
Sll
—
Marliave. 1975; Richard-
son, 1981a
2.9-3.0
"
>I6
Misitano, 1980; Richard-
son and Washington,
1980
3.1-3.3
5.5-7.3
12.6-15.0
Washington, 1981; Wash-
mgton, pers. obs.
4.2-4.5
—
—
Budd. 1940; Washington,
1981
-4.0
6.4-9.6
13-14
Washington. 1981
5.1-5.4
6.2-8.1
12.9-13.5
Washington. 1981
4.6-4.7
_
9-11
Moms. 1951
5.5-6.3
-7.0
—
Stein, 1972; Richardson
and Washmgton, 1980
6.3-6.9
—
9-10
Heufelder, 1982
6.86
—
9.5-10
Wallus and Granneman,
1979
5.7-6.3
—
8-11
Wallus and Granneman.
1979
10.5
_
_
Watanabe, 1976
—
—
—
Watanabe, 1976
WASHINGTON ET AL.: SCORPAENIFORMES
409
Table 107. Continued.
Family ''subfamily
species
Type of egg
pelagic (P).
demersal
(D) or vivip-
arous (V)
Number Largest oil
Egg size of oil globule size
(mm) globules (mm)
Body length (mm) at
Transfor-
Hatching
Flexion mation
References
7-8
Heufelder. 1982
4.9-5.2
5.2-7.0 7.6-
-7.8
Misitano, 1978; Richard-
son and Washington.
1980
5.5-5.8
12
Russell, 1976
-4.0
_ _
Russell, 1976
5.6-6.1
— —
Kyushin, 1970
C, riceP
Enophrys bison
Rhamphocottus richard-
soni
Tnglops murrayP
T. pingelP
Cottocomephoridae
Abyssocottus bergianus'
A. godlewskiP
A. korolnejjp
D
D
E. bubalis ( Taurulus) D
E. HUjeborgi ( TaurulusY D
Gymnocanthus herzen- D
steini
G. venlralis^ —
Hemileptdotus gitbertP —
H. hemilepidons D
H. jordani —
H. spinosus —
H. zapus —
Hemilriplems americanus D
H. villosus D
Iceliis btcornis^ —
Leptocottus armatus' D
Myoxocephalus aenaeus D
M. octodecimspinosus D
M. guadricornus D
M. scorpius D
M. thompsonP —
Naulichlhys oculofascia- D
tus
Oligocottus maculosus D
O. snyderi
Onhonopias triads'
Paricehnus hoplilicus^
Pseudobiennius cotloides^ D
Radulinus asprellus —
R. boleoides^ —
D
Scorpaenichlhys marmor- D
atus
D
1.7-1.8
1.5-1.8 several
2.0 1
1.6-1.7 few
1.5-1.6 1
2-2.5
0.36
0.38
0.31-0.56 -5-6
-5
-9.1
1 large
0.8
D
1.3-1.5
1 large
many
small
D
1.2-1.3
-
D
0.9-1.0
1 large
2 small
D
3.2-3.3'
—
D
2.8-3.0'
_
D
-4.5'
_
> 19-23
7.6-10.1 19
- 7-12 -
10-14 -14.5 >18.(
10.9-11.6 S14.4 -20
1.4-1.5
present
—
3.9-4.8
-8
15-20
1.5-1.7
2-1-
0.2
4.7-6.3
6.8
-
1.9-2.3
1 or more
diameter
varies
6.3-7.3
9-11
-15
1.5-2.2'
_
_
-8
ca. 10.5-1-
_
1.8-2.5
several
0.4-0.5
7.4-8.6
9-15
17-20
8-10 - -
9 -9-11 -26
4.2-4.5 7.2-7.6 7.5-10
4.47
6.2-8.4
11-13
2.9-3.8
-
-
<5.6
-
-25
12.5
£4.7
7.2-10.9
al5
-
-8.7
-
6-7
8.4
-14-15
2.0-2.2
2.5-2.8
1.4-1.9 1 large 0.27 5.8-6.0 7.5-8.7 14-15-
several
small
- - - 7-8 12 -
2.0 manv _ _ _ _
Ehrenbaum, 1905-1909
Hattori, 1964
Richardson and Washing-
ton, 1980
Gorbunova, 1964a''
Richardson and Washing-
ton, 1980
Matarese and Vinter (in
prep.)
Fahay, 1983; Fuiman,
1976
Okiyama and Sando,
1976
Russell, 1976; Ehren-
baum, 1905-1909
Richardson and Washing-
ton, 1980; Jones, 1962
Fahay. 1983; Lund and
Marcy, 1975
Fahay, 1983; Colton and
Marak, 1969
Khan and Faber. 1974
Russell. 1976; Ehren-
baum, 1905-1909;
Mcintosh and Master-
man, 1897
Heufelder, 1982
Richardson and Washing-
ton, 1980; Marliave,
1975
Washmgton, 1981; Stein,
1973
Washington, 1981; Stein,
1972
Bolin, 1941
Richardson and Washing-
ton, 1980
Watanabe, 1976
Richardson and Washing-
ton, 1980
Richardson and Washing-
ton, 1980
Richardson and Washing-
ton. 1980; Marliave,
1975; Blackburn, 1973
Richardson and Washing-
ton, 1980; O'Connell,
1953
Fahay, 1983
Bigelow and Schroeder,
1953; Rass, 1949
Taliev, 1955
Taliev, 1955
Taliev, 1955
410
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 107. Continued.
Type of egg
pelagic (P).
Jody length (mm
at
demersal
Number
Largest oil
Family/subfamily
(D) or vivip-
Egg size
of oil
globule size
Transfor-
species
arous (V)
(mm)
globules
(mm)
Halching
Flexion
mation
References
A. pallidus'
D
2.6-2.8'
—
_
-6
—
-16
Taliev, 1955
Asprocottus gibbosus^
D
3.3-3.4'
—
—
—
—
—
Taliev. 1955
A. herzensteiniP
D
3.0-3.2'
—
—
—
—
a9
Taliev, 1955
A. megalops'
D
3.5-3.7'
—
—
—
—
—
Taliev, 1955
Batrachocottus baicalensis
D
-3.0
3-10
small
10.0
post-
flexion'
N/A
-16
Chemyayev, 1981
B. multiradialus^
-4.0
—
—
-6.0
—
—
Taliev, 1955
B. nikolskii^
D
2.9-3.1
—
—
_
_
—
Taliev, 1955
B. uschkanP
—
—
—
—
—
—
_
Taliev, 1955
Cottinella boulengerp
D
2.8'
—
—
—
—
—
Taliev, 1955
Cottocomephorus gre-
D
1.2-1.8
—
—
-6.8-7.0
_
-19
Taliev, 1955
wingki
C. inermis^
D
1.5-1.7
—
—
—
_
_
Taliev, 1955
Paracottus kessleri
D
1.0-1.45
1 large
0.3
5.2-5.4
-6.2
-20
Chemyayev, 1978
P. kneri
D
2.0-2.3
_
_
6.8-7.1
_
>10.8
Taliev, 1955
Procotlus jeittelesP
D
2.5-3.3
-
-
-
-
-
Taliev, 1955
Hexagrammidae
Hexagrammos agrammys
D
2.02-2.07
many co-
alesce
to 1
8.15-8.61
-11
S 40-48
Fukuhara, 1971
H. decagrammus
D
—
—
—
ca. 8
15-18
-30"
Kendall and Vinter, 1984
H. lagocephalus
D
2.0-2.6
many
-8-9
12-15
ca. 29
Kendall and Vinter,
1984; Gorbunova,
1964b (as H. decagram-
mas)
H. octogrammus
D
1.75-2.10
many
0.8
6-7
-12-15
-30"
Gorbunova, 1964b
H. otakii
D
2.3-2.7
many
—
6.5-7.0
-11
—
Gorbunova, 1964b; Yusa,
1960c
H. stelleri
D
—
—
—
-7-9
-12-15
-30''
Kendall and Vinter, 1984
Ophiodon elongalus
D
2.9-3.2
1
—
-9.0
11-15
-30-
Kendall and Vinter, 1984
Pleurogrammus mono-
D
2.1-2.8
many
1.38-1.4
10-11
-14-19
-30"
Yusa, 1967; Gorbunova,
pterygius
1964b
Oxylebius pictus
D
—
—
—
4-5
7-9
-45
Kendall and Vinter,
1984; DeMartini, 1976
Zaniolepis sp.
—
—
—
—
-2.5
~6
?15
Kendall and Vinter, 1 984
Normanichthyidae
Normanichthys crockeri
-
-
—
—
<4.4
7-9
>16
Balbontin and Perez,
1980
Psychrolutidae
Dasycoltus seliger'
—
—
—
—
—
-10
—
Richardson, 1981a
Gilbertidia sigalutes-
D
2.3
—
—
—
-13-15
-23
Marliave, 1975
Malacocottus sp.
—
—
—
—
—
-7-9.8
£24
Richardson, 1981a
Psychrolutes paradoxus^
-
-
-
-
—
-10.5
-13-14
Marliave, 1975
' Ovarian or newborn larvae of 30 species of Sebasres not listed here are described in Efremenko and Lisovenko (1970). Westrheim (1975), and Moser el al. (1977).
^ Incomplete descnption with illustration.
' Rifle ovarian egg diameter.
• Pelagic juvenile stage.
' Hatch at advanced postflexion stage.
' Confusion exists regarding correct identification [Matarese and Vinter (in prep.)].
and Moser et al. (1977) on scorpaenids and Sparta (1956b) and
Richards (in prep.) on triglids and peristediids.
Scorpaenidae (Figs. 220-223).— This is the largest and most
diverse scorpaenoid family with about 44 genera and more than
350 species. The classification and relationships of the family
are in controversy (Washington et al., this volume) and we
follow their subfamily groupings.
Sebastinae. — Barsukov (1981) includes 3 genera and 1 1 4 species
in this temperate and boreal group. Sebastes with about 106
species accounts for almost '/3 of the species in the order. At
least a single larval stage is known for 62 species of Sebastes
and flexion or postflexion stages have been described for about
32 of these (Table 107). Larval stages have been described for
one of the 6 species oi Helicolenus and are unknown for the two
species of Hozukius.
In Sebastes most of the yolk is utilized before hatching while
the eggs lie freely within the ovary. Hatching precedes extrusion
and newborn larvae range from 3.8 to 7.5 mm in length among
the various species and have functional eyes, jaws, and pectoral
fins. The finfold is slightly inflated and has minute cell-like
WASHINGTON ET AL.: SCORPAENIFORMES
411
LIO
APO-4^
pp6-4
PPO-3
— LOP
B
TM
UIO-4
PPO-5
LIO-I
LIO-2
— PPO-3
PPO-5
APO-4
PPO-4
Fig. 220. Head spines in 6.2 mm (A), 8.2 mm (B). 10.0 mm (C) and 16.0 mm (D) stained larvae of Sehasles melanostomus. Abbreviations
of head spines: APO-2, 2nd anterior preopercular; APO-3, 3rd anterior preopercular; APO-4, 4th anterior preopercular; CL, cleithral; lOP,
interopercular; LIO-1, 1st lower infraorbital; LlO-2, 2nd lower infraorbital; LOP, lower opercular; LPST, lower posttemporal; NA, nasal; NU,
nuchal; PA, parietal; PPO-1, 1st posterior preopercular; PPO-2. 2nd posterior preopercular; PPO-3, 3rd posterior preopercular; PPO-4, 4th
posterior preopercular; PPO-5, 5th posterior preopercular; PRO, preocular; PSO, postocular; PT, pterotic; SC, supracleithral; SPO, supraocular;
TM, tympanic; UIO-1, 1st upper infraorbital; UIO-2, 2nd upper infraorbital; UIO-3. 3rd upper infraorbital; UIO-4, 4th upper infraorbital; UOP,
upper opercular; UPST, upper posttemporal. From Moser and Ahlstrom, 1978.
Structures concentrated along the dorsal and ventral margins.
Notochord flexion occurs at about 6-12 mm and transformation
at 15-25 mm (Table 107). Many species have a distinct pelagic
juvenile stage which can reach almost 60 mm body length.
Preflexion larvae have a slender body (body depth 1 3-23%
of body length) and compact gut; snout-anus distance increases
from about 40-50% of body length to over 60% in some species
during the larval period. The caudal and pectoral fins begin
forming first, followed by the pelvics and then the dorsal and
anal fins. The pectoral fins range from short and rounded to
elongate and fan-shaped, reaching almost 50% of body length
in 5. levis (Fig. 221). The pectoral fin base is shallow (typically
7-13% of body length) in comparison with other subfamilies.
Ossification of skeletal elements begins early in the larval period
and proceeds rapidly as in other scorpaenoids; vertebral ossi-
fication follows the pattern of other scorpaeniforms, with the
neural arches ossifying before the centra (Moser, 1972).
Pigmentation in newborn larvae consists of a melanistic sheath
over the gut and a postanal series along the ventral midline.
Some species also have a dorsal midline series which may de-
velop gradually. Pigment increases with development, appear-
ing on the head (above brain, on jaws and opercular region),
fins, and caudal peduncle. Often the pectoral fins (both base and
blade) have diagnostic pigment patterns. Several of the western
Pacific species are heavily pigmented with the head and body
covered by a sheath of melanophores (Fig. 221).
Head spines are a prominent feature of all Sebastes larvae.
Pterotics, parietals (usually serrated), and preopercular spines
412
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 221. Larvae of Scorpaenidae. (A) Sehastes oblongiis. 8.5 mm TL (from Fujita, 1958); (B) 5. longispinis. 7.1 mm TL (from Takai and
Fukunaga, 1971); (C) S. huhbsi. 6.0 mm TL (from Uchida et al., 1958); (D) S. zacentrus. 12.7 mm SL (from Laroche and Richardson. 1981); (E)
5. paucispims. 10.5 mm SL (from Moser et al., 1977); (F) 5. jordani. 15.5 mm SL (ibid.); (G) S. levis. 10.4 mm SL (ibid.); (H) Hetwolenus
daclyloplerus, 10.0 mm (from Tuning, 1961).
form during the preflexion period in most species, and other
spines appear gradually thereafter (Fig. 220). Although there is
variation in larval spine complements (Moser and Ahlstrom,
1978; Moser and Butler, 1981; Richardson and Laroche, 1979;
Laroche and Richardson. 1980, 1981). it is apparent that 1)
the adult head spine complement develops during the larval
period and 2) certain spines develop during the larval period
but are not present in adults. Of the latter, the most prominent
are the pterotic, anterior preoperculars, lower posttemporal, and
upper infraorbitals.' The fact that these spines do occur in adults
of other subfamilies is of possible phylogenetic significance
(Moser and Ahlstrom, 1978).
Upper infraorbitals are present in adults of a few species ofSebastes.
Helicolenus is viviparous, the fertilized eggs developing in a
gelatinous matrix within the ovary (Graham, 1939; Krefft, 1961).
Larvae of H. dactyloptenis have been described; hatching and
birth occur at a smaller size (2.2 mm) than in Sebastes. although
sizes at notochord flexion and transformation are similar (Table
1 07). Larvae are moderately deep-bodied (Fig. 221); body depth
averages 29%, 33%, and 49% of body length for preflexion,
flexion and postflexion stages. Head and gut shape are similar
to that of Sebastes. The pectoral fin is moderate in size and
rounded; the base is slightly deeper than in most species of
Sebastes. Sequence of fin formation is similar to that oi Sebastes.
A mass of spongy tissue develops anteriorly in the dorsal finfold
in preflexion larvae and persists through most of the larval
period; the structure is apparently unique. The early pigment
pattern consists of a dorsolateral gut sheath, melanophores above
the brain, on the lower jaw, in a short median ventral series just
WASHINGTON ET AL.: SCORPAENIFORMES
413
Fig. 222. Larvae of Scorpaenidae. (A) Ponlmus Type A. 8.0 mm SL (from Moser et al., 1977); (B) Scorpaena Type A, 8.0 mm SL (ibid.); (C)
Scorpaenodes xyris, 6.2 mm SL (ibid.); (D) Sebastotobus sp. 7.7 mm SL (ibid.); (E) Ectreprosebasies imus. (s.l mm SL (ibid.).
anterior to the caudal fin, and on the distal and proximal regions
of the pectoral fin blade (Fig. 221). Head spine formation is
similar to that of Sebastes species which have full larval com-
plements, except that spines are lacking on the 2nd infraorbital
bone and the cleithrum.
Scorpaeninae. — Larval stages are known for only 3 of the 15
genera in this subfamily; a total of 8 species (or generic types)
out of about 150 have been described (Table 107; see Sparta,
1956b and Moser et al., 1977, for major reviews). Hatching
occurs at about 2.0 mm or less; newly-hatched larvae have a
large elliptical yolk sac, unpigmented eyes, pectoral fin buds,
and lack a mouth. The finfold is inflated and, along with the
body skin, forms a balloon-like envelope that is attached prin-
cipally at the snout and pectoral regions (Orton, 1955d). Cell-
like granulations cover the entire envelope but are concentrated
414
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 223. Larvae of Minoinae (A). Triglidae (B, C), Congiopodidae (D, E), Platycephalidae (F), Hoplichthyidae (G, H). (A) Minous sp.?, 6.4
mm SL (from Leis and Rennis, 1983); (B) Prionolus sp., 6.4 mm SL (original); (C) Prionotus slephanophrys. 8.8 mm SL (CalCOFI 7510 sla.
1 1 7.70); (D) Congiopodus spimfer. 1 0.8 mm SL (from Brownell, 1 979); (E) Detail of pectoral fin of Congiopodus spmifer (ibid.); (F) Platycephalidae,
unidentified, 6.2 mm SL (from Leis and Rennis, 1983); (G) Hoplichthys sp., 7.1 mm SL (original, courtesy M. Okiyama); (H) Hoplichlhys sp.
17.2 mm SL (ibid.).
at the median edges of the finfold. Flexion occurs at a small size
(4-6 mm) as does transformation (10-17 mm). Larvae are rel-
atively deep-bodied during preflexion and flexion and more so
during postflexion, when body depth averages 38-40% of body
length for the genera listed in Table 107. The gut is compact
and the head becomes massive. Snout-anus length increases
from 46-50% of body length in preflexion larvae to 61-67% in
postflexion larvae. The snout has a steep profile (Fig. 222).
The pectoral fins are well developed and deep-based; fin base
depth is 13-15% of body length in preflexion larvae and 14-
1 8% in flexion and postflexion larvae. They are fan-shaped and
enlarged in Scorpaenodes; fin length attains 41% of body length
during the postflexion stage. They are smaller but distinctively
shaped in Scorpaena (fan-shaped with scalloped margin) and
Pontinus (slightly wing-shaped). Ossification of fin rays, as well
as skeletal elements, occurs in early larvae (4-5 mm). The cau-
dal, pectoral, and pelvic rays begin ossifying almost simulta-
neously, followed immediately by the dorsal and anal fins.
Preflexion larvae have a postanal ventral midline series of
melanophores ranging in number from 2-7 in Scorpaena guttata
to 12-18 in Scorpaenodes xyris. The most prominent pigment
is on the pectoral fins; typical patterns are a concentration at
the distal margin (Scorpaenodes, some Pontinus, some Scor-
paena spp.), a solid covering over most of the fin (some Scor-
paena spp.), or a diagonal bar (some Pontinus spp.). A melanistic
sheath develops over the dorsal surface of the gut and gas bladder
in most species of Scorpaena. whereas in Scorpaenodes and
Pontinus only the gas bladder is pigmented. Other pigment in
Scorpaena forms at the cleithral juncture and above the brain
(Fig. 222).
Fig. 224. Larvae of the Oxylebius scorpaeniform group (A, B) and the hexagrammid group (C-F) of Washington and Richardson (MS) (see
Washington et al., this volume). (A) Oxylebius pictus. 8.5 mm SL (from Kendall and Vinter, 1984); (B) Zaniolepis sp., 7.7 mm SL (ibid.); (C)
Hexagrammos oclogrammus, 15.2 mm SL (ibid.); (D) Pleurogrammus monopterygius, 20.5 mm SL (ibid.); (E) Ophwdon elongalus, 1 5.4 mm SL
(ibid.); (F) Anoplopoma fimbria, 13.8 mm SL (Ahlstrom and Stevens. 1976).
WASHINGTON ET AL.: SCORPAENIFORMES
415
416
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
B
:^\V\\\v'^\v\\\\\ \V\\ \ \
Fig. 225. Larvae of Normanichthyidae (A), Cottocomephoridae (B, C), Comephoridae (D, E). (A) Normanichlhys crockeri. 8.5 mm SL (original);
(B) Coltocomephonis grewingki. 7.4 mm (from Taliev, 1955); (C) Cottocomephorus inenrus. 1 1.2 mm (ibid.); (D) Comephorus baicalensis. 6.9
mm (ibid.); (E) Comephorus baicalensis. 21.3 mm (ibid.).
Cranial spine development is similar to that in sebastines.
The pterotic, parietal, postocular (supraocular crest), posterior
preoperculars (2nd, 3rd, and 4th) anterior preoperculars (2nd
and 4th) and lower posttemporal develop during the prefiexion
period. The lower infraorbital ( 1 st), upper infraorbitals ( 1 st and
4th), posterior preoperculars (1st and 5th), nuchal, supraclei-
thral, cleithral, upper opercular, and lower opercular spines ap-
pear during postflexion. Late in the postflexion stage the lower
infraorbital (2nd), nasal, preocular, and supraocular spines ap-
pear. Spines which do not develop in scorpaenine larvae but
are present in adults of most genera are the upper infraorbitals
(2nd and 3rd), upper posttemporal, tympanic, and sphenotic.
In Scorpaenodes the nuchal spine develops during the prefiexion
period and exceeds the parietal spine in length, giving the pa-
rietal ridge a bifurcate appearance. In other scorpaenines and
all other scorpaenids except Sebastolobus. the nuchal develops
late and is excluded from the parietal ridge.
Pteroinae.— Early prefiexion larvae have been described for
Pterois lunulata and Dendrochirus brachypterus (Table 107).
Newly-hatched larvae are small (1.1-1.6 mm) and similar in
morphology to those of Scorpaeninae. The pectoral fins are large
and fan-shaped with pigment at the distal margin. Postanal
pigment in Pterois consists of ventral and dorsal midline series.
In Dendrochirus this pigment coalesces to form a band.
Sebastolobinae. — Life history series have been described for Se-
bastolobus alascanus and S. altivelis (Moser, 1974). Larvae are
2.6 mm at hatching, 6.0-7.3 mm at notochord flexion, and 14-
20 mm at transformation. The distinctive pelagic juveniles (up
to 56 mm in S. altivelis) have a prolonged midwater existence
before settling to the deep shelf and slope habitat of the adults.
Larval morphology is similar to that of scorpaenines. The pec-
toral fins are large, deep-based, and fan-shaped (Fig. 222); their
rays are the first to ossify, followed by the caudal rays and then
WASHINGTON ET AL.: SCORPAENIFORMES
417
■r,i',h: •.
>j«-.i;
dft^/
Fig. 226. Larvae of the Rhamphocottus group (A) and the Scorpaenichthys group (B. C) of cottids of Washington and Richardson (MS) (see
Washington et al., this volume). (A) Rhamphocottus nchardsoni. 10.6 mm SL (from Richardson and Washington, 1980); (B) Scorpaenichthys
inarmoratus. 8.7 mm SL (ibid.); (C) Hemilepidotus spinosus, 1 1.0 mm SL (ibid.).
those of the other fins. The pectoral fins are pigmented at the
distal margin; other pigment includes a sheath over the gut and
melanophores above the brain. Head spination is highly de-
veloped (Fig. 222); the sequence of development is similar to
that of scorpaenines. In addition to the spine complement of
scorpaenines, Sebastolobus larvae develop the 2nd and 3rd up-
per infraorbital spines and the 1 st anterior preopercular spine.
Setarchinae. — Larvae are known for Ectreposebastes imus (Moser
et al., 1977). Hatching and notochord flexion occur at a small
size as in the scorpaenines; however, postflexion larvae attain
a large size (Table 107). Larvae have the deepest body of known
scorpaenids; body depth reaches 55% of body length in late
postflexion stage. The gut is compact with an elongate terminal
section; snout-anus distance averages 53% of body length in
preflexion larvae and 76% in postflexion. The pectoral fins are
deep-based, fan-shaped, and large, extending to the caudal pe-
duncle (Fig. 222). Fin base depth and fin length reach 22% and
57% of the body length respectively. The pigment pattern con-
sists of a postanal ventral series of 1 1-14 melanophores (not
present after 4.0 mm), a blotch above the gas bladder, and an
almost solid sheath over the pectoral fin, which recedes distally
with development. Head spine development is similar to that
of scorpaenines.
4^:^^€:^^^^^&>^
WASHINGTON ET AL.: SCORPAENIFORMES
419
Fig. 227. Larvae of the Myoxocephalus group of cottids of Washington and Richardson (MS) (see Washington et al., this volume). (A)
Paricelinus hopliticus. 13.8 mm SL (from Richardson and Washington, 1980); (B) Triglops sp.. 15.4 mm SL (ibid.); (C) Icelus hicornis. 25 mm
(from Ehrenbaum, 1905-1909); (D) Chitonotus pugetensis. 11.5 mm SL (from Richardson and Washington, 1980); (E) Artednis meanyi. 13.8
mm SL (ibid., as Iceiinus sp.); (F) Icelinus sp., 1 1.9 mm SL (original); (G) Ascelichthvs rhodorus. 1 1.0 mm SL (from Matarese and Marliave,
1982).
Choridactylinae.— The developmental stages of Ininucus ja-
ponicus have been described by Fujita and Nakahara (1955) and
Sha et al. (1981). Larvae are 3.2 mm at hatching, 6.4-8.2 mm
at flexion and about 10 mm at transformation. Yolk-sac larvae
are similar to those of Scorpaeninae. Larvae are relatively slen-
der and blunt-headed, with a compact short gut (Fig. 223). The
pectoral fins are large and fan-shaped, with a scalloped margin;
they develop a series of large blotches distally. One to several
large postanal melanistic blotches form on the postanal ventral
midline and the gas bladder region is pigmented. Sha et al. ( 1 98 1 )
show the larvae to be heavily xanthic.
Minoinae. — Leis and Rennis (1983) described a larval series
tentatively identified as Minous sp. It- is generally similar in
morphology and pigmentation to Jnimicus; however, the pec-
toral fin is relatively larger and has a different pigment pattern.
Triglidae (Fig. 223). — E^s are only known for 3 of the 8 genera
of triglids. The new world genus Prionotus has multiple oil
globules whereas single oil globules are known for Chelidonich-
thys and Lepidotrigla. Larvae are poorly known with complete
series having been described for 4 species in 3 genera (Table
107). There are approximately 90 species in this family and
many are very difficult to identify as adults. The genus Lepi-
dotrigla has 40+ species and is poorly known in many areas.
Diagnostic features include the depressed profile of the head and
large pectoral fins of which the lowest three rays become de-
tached during transformation. Meristics are very similar to
platycephalids and caution is advised. However, most triglids
have fewer pectoral rays than most scorpaenoids. Prionotus.
including Bellator, has 1 3 to 15 plus 3 free rays; Trigla, Chel-
idonichthys. Lepidotrigla, and Uradia have 11 plus 3 free rays;
and Pterygotrigla and Parapterygotrigla have 11 to 13 plus 3.
Peristediidae. — ELH information has been published only for
Peristedion cataphractum of the eastern Atlantic (Table 107).
Larvae and transforming juveniles have elongated upper pec-
toral rays and strong head spination (see plate 40 in Padoa,
1956e). This family is often combined with the Triglidae, but
differs in many characters such as the presence of barbels, 2
rather than 3 free pectoral rays, and the body is encased in bony
scutes rather than scales. Three genera (Heminodus. Parahem-
inodus and Gargariscus) have jaw teeth and two genera (Per-
istedion and Satyrichthys) lack jaw teeth. There are about 25
species found in the tropics of all oceans in deep water (>200
m).
Congiopodidae (Fig. 223). — Eggs are known for only 1 {Con-
giopodus) of the 4 genera of Congiopodidae (Brownell, 1978;
Gilchrist, 1904; Robertson, 1974). The pelagic eggs are rela-
tively large (1.7-2.18 mm) and spherical, with a narrow peri-
vitelline space and no oil globules. The egg surface is covered
with striations. Early life history stages have been illustrated for
one species, Congiopodus spimfer{Qvov<mt\\, 1979; Gilchrist and
Hunter, 1919). Robertson (1975a), illustrated a well-developed
embryo of C. leucopaecilus. Larvae hatch at about 5 to 6 mm
NL and are elongate with long guts reaching 50% SL. The pec-
toral fins are extremely large and fan-shaped. Melanistic pig-
ment is present on the head, nape and on the dorsal and ventral
surface of the gut. Two large blotches of pigment on the dorsal
and ventral midlines form a band midway between the vent
and tail tip. The large pectoral fins have a distal band of pigment
which gradually expands over the entire fin with development.
Larvae develop large postocular and parietal spines. The pres-
ence of preopercular spines can not be determined from the
description by Brownell (1975).
420
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
\
-i«75f*:
WASHINGTON ET AL.: SCORPAENIFORMES
421
Platycephalidae (Fig. 22ij. — Platycephalids spawn small spher-
ical eggs (< 1 mm) with a single oil globule (Chang et al.. 1980;
Uchida et al., 1958). Larvae have been described and illustrated
for Platycephalus indicus (Ueno and Fujita, 1958) and for a
series of larvae incorporating seven unidentified species (Leis
and Rennis, 1983). Newly-hatched platycephalids are relatively
small (1.7-2. 3 mm) and slender-bodied, with unformed mouths,
unpigmented eyes, and large yolk sacs. By the time of yolk
absorption larvae have large heads and deep bodies which taper
toward the tail. The gut is quite long reaching % SL during
development. The pointed snout becomes distinctively long and
flattened. Pigmentation is usually present on the head, jaws,
ventral surface of the gut and along the postanal ventral midline.
Pigment may also be present on the dorsolateral surface of the
tail and pectoral fin. Larvae develop 4 to 9 preopercular spines.
Other head spines include: supraocular, supracleithral, parietal
and pterotic. Unlike most other scorpaeniforms, head spines
persist and become more pronounced in juveniles. Fin devel-
opment proceeds as follows: pectoral, caudal, dorsal, anal and
pelvic.
Hoplichthyidae (Fig. 22ij. — The pelagic eggs of Hoplichthys
haswelli are described by Robertson (1975a) as small and spher-
ical with a smooth surface. A single oil globule is present. De-
scriptions of hoplichthyid larvae have not been published; how-
ever, based on Okiyama (in prep.) larvae are quite similar
to platycephalids. Preflexion larvae (3.2 mm) are elongate with
large heads and pointed snouts. The gut is moderately long
(>50% SL) and the early-developing pectoral fins are large and
fan-shaped. The snout becomes increasingly long and depressed
during development. Pigmentation is limited to the gut, distal
tip of the pectoral fin and a band on the ventral finfold midway
between the vent and notochord tip. Numerous clusters of small
spines develop in the supraocular, parietal and pterotic regions.
Seven spines form on the posterior margin of the preopercle
with smaller spines at their base. As in platycephalids, head
spines persist in juveniles.
Dadylopteridae (Fig. 233).— The pelagic eggs are small (0.8 mm)
and slightly ovoid with a single oil globule. The egg surface is
smooth and unsculptured. Larvae hatch at about 1.8 mm and
undergo flexion of the notochord between 3.9-6.5 mm. Trans-
formation to the juvenile form occurs at about 9 mm. Larvae
are moderately deep-bodied with a distinctively blunt snout and
small mouth. The gut is long, reaching about 75% SL in post-
flexion larvae. Pigmentation occurs over the head, gut, along
the postanal ventral midline and around notochord tip. Pig-
mentation increases dramatically over most of the body in post-
flexion larvae. The distinctive head armature is quite different
from all known scorpaeniform larvae and is present in larvae
as small as 2.3 mm NL. A small supraoccipital spine is present
only during the larval period. The extremely long posttemporal
and preopercular spine extend posteriorly to the middle of the
anal fin in larvae by about 6.5 mm and persist in juveniles and
adults.
COTTOIDEI
Eggs
Eggs are known from representatives of six of the nine cottoid
families recognized here (Table 107). Where known, most cot-
toids spawn demersal, adhesive eggs which often form clusters
found under rocks. Eggs are frequently brightly colored, e.g.,
red, blue, green, yellow. The eggs of Anoplopoma fimbria are
pelagic. The Comephoridae of Lake Baikal are reported to be
viviparous.
Most eggs are spherical and average 1-2 mm in diameter,
although eggs as large as 4 mm have been reported in the cottid
Hemitripterus and some of the cottocomephorids. A single large
oil globule, frequently accompanied by several small ones, oc-
curs in many species. The surface of the eggs is often covered
by a tough adhesive membrane, and may be smooth as in An-
oplopoma and Myo.xocephalns aenaeus (Fahay, 1983) or cov-
ered by tiny, radiating canals as in Arledius lateralis and Cli-
nocottus analis (Budd, 1940).
Larvae
At least one larval stage is known for 88 of the 329+ species
and for 46 of the 104 genera of cottoids. Major overviews of
larval cottoid taxonomy include: Richardson and Washington
(1980) on cottids; Kendall and Vinter (1984) on hexagrammids;
Taliev (1955) and Chemyayev (1971, 1975, 1978, 1981) on
comephorids and cottocomephorids; and, forthcoming Laroche
(in prep.) on agonids.
Larval cottoids exhibit a broad diversity of form. Size at
hatching varies from 2 to 12 mm. Planktonic life may be quite
brief several weeks in many cottids, or may be extended up to
a year with a special pelagic juvenile stage as in the hexagram-
mids.
Cottoid larvae exhibit such a diversity of form and devel-
opment that it is impossible to characterize a generalized "cot-
toid" larva.
Hexagramrnidae (Fig. 224). — Larvae are known for 10 of the
1 1 species of the hexagrammid genera Hexagrammos. Pleuro-
grammus, and Ophiodon. Major works presenting descriptions
and illustrations include Kendall and Vinter (1984) and Gor-
bunova (1964b). Hexagrammids hatch at a relatively large size
(6-1 1 mm NL). Development is gradual from hatching to the
juvenile stage with a prolonged epipelagic prejuvenile period
(~ 30-50 mm SL). Larvae have elongate, slender bodies with
large eyes. Larval Hexagrammos and Pleurogrammus have blunt
heads, while Ophiodon larvae have pointed snouts and large
terminal mouths.
Larvae are heavily pigmented especially dorsally. Melano-
phores are scattered over the head, gut and usually on the dorsal
and ventral midlines. The extent of postanal, ventral midline
and lateral pigmentation is useful in specific identification.
Fin formation proceeds in the following sequence: caudal,
pectoral, second dorsal and anal, first dorsal and pelvic. Larvae
exhibit delayed ossification. Vertebral ossification in hexagram-
Fig. 228. Larvae of the Myoxocephalus cottid group of Washington and Richardson (MS) (see Washington et al., this volume). (A) Onhonopias
triacis. 7.0 mm SL (original); (B) Enophrys bison. 7.0 mm SL (from Richardson and Washington, 1980); (C) Myo.xocephalus aenaeus, 7.0 mm
SL (from Lund and Marcy, 1975); (D) Myo.xocephalus polyacanlhocephalus. 12.0 mm SL (from Richardson, 1981a); (E) Radulinus asprellus,
10.9 mm SL (from Richardson and Washington, 1980); (F) Gymnocanthus tncuspis. 13.0 mm (from Khan, 1972).
422
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
mids (and Anoplopoma) is similar to that in Scorpaenoidei with
the neural and hemal arches ossifying before the associated ver-
tebral centra. Vertebral counts are notably high (47-63). Head
spines are absent in larval Hexagrammos and Pleurogrammus
and extremely reduced in Ophiodon. with late-stage larvae de-
veloping 4 tiny preopercular spines.
Anoplopomatidae (Fig. 224. Table ;07j. -Larvae of only Ano-
plopoma have been described and illustrated by Kobayashi (1957)
and Ahlstrom and Stevens (1976). Early development oi Ano-
plopoma is similar to that of the hexagrammids. Larvae hatch
at a large size (~ 9 mm NL) and development is gradual without
great changes in form.
Larvae are slender and elongate with pointed snouts and long
guts. The distinctive pectoral fins with heavy distal pigmentation
are exceptionally large reaching nearly 33% SL late in the larval
period. Larvae are heavily pigmented with melanophores over
most of the head, gut and lateral surface of the body.
As in hexagrammid larvae, ossification is delayed with the
neural and hemal arches ossifying before the associated vertebral
centra. Vertebral counts (61-66) are distinctively high. Pectoral
fin development is precocious. Head and preopercular spines
are absent.
Oxylebius-Zaniolepis (Fig. 224).- Oxylebius and Zaniolepis are
sometimes included in the Hexagrammidae, but are herein treat-
ed separately because of the distinctiveness of their larvae from
hexagrammids (Washington and Richardson, MS; Kendall and
Vinter, 1984). Larvae of Oxylebius pict us and Zaniolepis sp. are
illustrated and described by Kendall and Vinter (1984). Larvae
hatch at a small size (2.5-5 mm NL), undergo notochord flexion
between 6 and 9 mm NL, and transform to a benthic juvenile
at about 1 5 mm SL.
Oxylebius and Zaniolepis are relatively short and deep-bodied
with large, bulging guts and rounded snouts. Pectoral fins de-
velop early and are distinctively large and fan-shaped. Pigmen-
tation is heavy over the anterior half of the body in preflexion
larvae and increases over the postanal lateral body witb devel-
opment. Zaniolepis possesses characteristic snout pigment which
is absent in Oxylebius. The pectoral fins of both species are
densely pigmented.
Head spination is well-developed with preopercular (5 spines
in Oxylebius; 6-7 in Zaniolepis), posttemporal and supraclei-
thral spines present. Zaniolepis larvae develop distinctive prick-
le-scales over most of the body by about 7 mm.
Normanichthyidae (Fig. 225^. -Larvae of the monotypic Nor-
manichthys crockeri are illustrated and described by Balbontin
and Perez (1980). Hatching occurs at a small size (4.4 mm NL)
and flexion of the notochord occurs at 7 to 9 mm. Development
from hatching to the juvenile stage is gradual without great
change.
Larvae are elongate and slender with short, coiled guts and
distinctive large pectoral fins. Pigmentation is restricted to the
pectoral fins and the ventral midline extending from the isthmus
to the tail. In small larvae several large melanophores are pre-
sent on the dorsal midline.
Distinctive features of larval development include: the ab-
sence of head and preopercular spines, delayed ossification, early
development of the pectoral fin, and presence of only 5 bran-
chiostegal rays.
Comephoridae (Fig. 225).— The endemic comephorids of Lake
Baikal in Russia are reported to be viviparous (Chemyayev,
1971, 1975) and are bom at a relatively large size (8.2-9.4 mm)
but are not well developed. Flexion of the notochord occurs at
about 8.2 to 13 mm. Larvae develop very slowly with trans-
formation occurring 3 or 4 months after birth.
Larvae are extremely slender and elongate with small heads
and very short coiled guts. Comephorids are quite different from
other cottoids morphologically and are blennioid in appearance.
Pigmentation is usually limited to the gut and sometimes in a
series along the postanal lateral midline. Four small preoper-
cular spines develop in late-stage larvae; other head spines are
absent.
Cottocomephoridae (Fig. 225). — Larvae of seven genera of Lake
Baikal cottocomephorids have been described and illustrated
(Chemyayev, 1971, 1975. 1978, 1981; Taliev, 1955). Larvae
hatch at about 5 to 10 mm, and range from forms with large
yolk sacs and no fin development (e.g.. Paracottus) to well de-
veloped, postflexion forms with fins well developed (e.g., Ba-
trachocottus). Size at transformation varies from 9 to 20 mm.
Larvae are slender with moderately short guts and rounded
snouts, somewhat similar to freshwater cotlids (Coitus) in form.
Pigmentation is variable with melanistic pigmentation usually
present on the head, nape, gut and variously on the dorsal and
ventral midline. Melanophores are frequently present in a row
along the lateral midline near the tail tip.
Larvae develop 4 small preopercular spines accompanied by
two spiny projections from an inner preopercular shelf Other
head spines are lacking.
Cottidae (Figs. 226-231). — The taxonomic status of the family
Cottidae is controversial with the number of recognized families
ranging from 1 to 17 (see Washington and Richardson, MS). To
minimize confusion, and because there is no generally agreed
upon classification of this "family," we use the generic groupings
identified by Washington and Richardson (MS) for our discus-
sion of early life history information. Larvae are known for 28
of the 70+ cottid genera. A general overview of larval cottid
taxonomy is presented in Richardson and Washington (1980),
Richardson (1981a), Washington (1981) and Fahay (1983).
Rhamphocottus (Fig. 226). — Larvae of this distinctive, mono-
typic species hatch at a relatively large size (6-7 mm NL). No-
tochord flexion occurs at 7 to 8 mm and transformation to a
Fig. 229. Larvae of the Artedius Part A group (A-C) and the Couus group of cottids of Washington and Richardson (MS) (see Washington
et al., this volume). (A) .Artedius fenestralis. 9.9 mm SL (from Richardson and Washington, 1980. as .Artedius Type 2); (B) ClmocoUus acuticeps.
10.4 mm SL (from Washington, in prep.); (C) Oligocottus snyderi. 10.2 mm SL (from Washington, 1981); (D) Leptocottus armatus. 8.1 mm SL
(from Richardson and Washington, 1980); (E) Cottus asper. 8.2 mm SL (ibid.).
WASHINGTON ET AL.: SCORPAENIFORMES
423
-.•I'V-^
y
%C:jk!ddi:::^:k^:::k:^::^^^
424
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 230. Larvae of the Psychro/ules group (A, B) and the Malacocollus group (C, D) of cottids of Washington and Richardson (MS) (see
Washington et al.. this volume). (A) Psychrolutes paradoxus. 13.0 mm SL (from Richardson. 1981a); (B) Gilbenidia sigalutes, 13.0 mm SL (ibid.);
(Q Dasycottus seliger, 10.3 mm SL (original); (D) Malacocollus zonurus, 9.8 mm SL (original).
WASHINGTON ET AL.: SCORPAENIFORMES
425
S5^
Fig. 231. Larvae of the Hemitripterus group (A-C) of cottids of Washington and Richardson (MS) (see Washington et al.. this volume) and
Agonidae. (A) Hemitripterus vitlosus. ca. 15.5 mm SL (from Kyushm, 1968); (B) Blepsias arrlwsus. 1 1.0 mm SL (from Richardson, 1981a); (C)
Nautichthys oculofasciatus. 1 1.7 mm SL (from Richardson and Washington. 1980); (D) Agonomalus or Hypsagonus sp., 8.2 mm SL (original,
courtesy B. Vinter).
benthic juvenile occurs at about 14 to 15 mm SL. Rhampho-
cottus larvae are extremely deep-bodied with a very long snout-
anus length.
Larvae are uniformly covered with melanophores except for
the caudal peduncle and ventral surface of the gut. Rhampho-
cottus develop small prickle-scales over most of the body by 9
or 10 mm. Larvae develop only one preopercular spine in con-
trast to the usual four possessed by most cottid larvae. Parietal,
nuchal, supracleithral, posttemporal and postocular spines occur
during the larval period.
Hemilepidotus-Scorpaenichthys (Fig. 226).— Larvae of this group
hatch at 4 to 6 mm NL. Transformation to the neustonic or
pelagic juvenile phase occurs at about 1 3 to 20 mm. Larvae are
Fig. 232. Larvae of Agonidae (all original). (A) Hypsagonus quadricornis. 11.5 mm SL; (B) Bolhragonus swani. 6.3 mm SL; (C) Xeneretmus
latifrons. 9.6 mm SL; (D) Slellerina xyosterna. 10.2 mm SL; (E) Ocella verrucosa, 10.1 mm SL; (F) Aspidophoroides monopterygius. 14.3 mm
SL.
WASHINGTON ET AL.: SCORPAENIFORMES
427
Fig. 233. Larvae of Dactylopteridae. (A) Daayloplerus voliians. 2.4 mm (from Padoa, 1956e); (B) Dactylopterus volitans. 7.5 mm (ibid.).
long and slender at hatching with moderately long guts (44 to
60%) and rounded snouts. They become increasingly deep-bod-
ied with development.
Larvae are relatively heavily pigmented with melanophores
over the head and gut. Scorpaenichthys larvae have dense pig-
ment covering the body except for the caudal peduncle while
Hemilepidotus spp. have postanal pigment concentrated on the
dorsal and ventral body midlines. Lateral melanophores de-
velop above and below the notochord in Hemilepidotus.
Hemilepidotus and Scorpaenichthys larvae develop four
prominent preopercular spines. Hemilepidotus possess numer-
ous head spines while Scorpaenichthys develop bony bumps in
corresponding areas. Larvae of this group develop unique pitted
dermal bones on the head. In addition, the uppermost pectoral
radial is tiny and fuses to the scapula in larval Scorpaenichthys
and nearly so in Hemilepidotus.
Myoxocephalus group (Figs. 227 and 228). — This is the least
well-defined and most diverse cottid group containing 1 3 genera.
Where known, size at hatching varies from 2.9 to 10 mm. Trans-
formation to the benthic juvenile stage vanes from 7.6 to 20
mm.
Members of this group are generally slender-bodied with
pointed snouts; however, Enophrys is stout-bodied, and Or-
thonopias has a blunt, rounded snout.
Pigmentation is variable. Heavy pigment on the dorsal surface
of the gut, on the nape and along postanal ventral midline is
characteristic of many members of this group. Several genera
possess heavy melanistic pigmentation on the lateral body sur-
face (e.g. Radulinus, some Myoxocephalus). Head pigment may
be present.
Larvae of this group develop four preopercular spines and a
distinctive bony preopercular shelf. Parietal, nuchal, supra-
cleithral, posttemporal and occasionally, postocular spines de-
velop in late-stage larvae.
Artedius group (Fig. 229, Table 107).— This group contains 3
genera, Artedius (in part), Clinocottus and Oligocottus and the
larvae have been described by Washington (1981). Larvae hatch
at 3 to 5 mm and transform to benthic juveniles at approxi-
mately 10 to 13 mm. Larvae are stubby-bodied with a slightly
humped appearance at the nape. Snouts are rounded and guts
trail distinctively below the ventral body midline. Several species
oi Artedius develop dorsal gut diverticula while Clinocottus acu-
ticeps develops long hindgut diverticula.
Larvae are relatively lightly pigmented and characterized by
pigment on the nape, over the gut and along the postanal ventral
midline. Head pigment is present in some species.
Larvae develop a unique preopercular spine pattern with 6
to 24 spines. Parietal and supracleithral spines are variable in
this group and may form in clusters, individually or not at all.
Leptocottus group (Fig. 229).— This group includes the genera
Leptocottus and Coitus. Hatching occurs at 4 to 5 mm and
transformation ranges from 8 to 12 mm. Larvae are relatively
slender-bodied with rounded snouts and moderately short guts.
Pigmentation is usually light with melanophores on the nape,
over the gut and widely spaced along the postanal ventral mid-
line. Head pigment may be present.
Where known, these larvae develop four weak preopercular
spines; however, other head spines are lacking.
Psychrolutes group (Fig. 230).— This group includes two genera
Psychrolutes and Gilbertidia. Larvae hatch at a relatively large
428
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
size, about 6 to 7 mm. They transform and settle from the
plankton at about 1 8 to 20 mm SL. Larvae are generally tadpole
shaped with large rounded heads tapering toward the tail. Larvae
possess an outer layer of loose flabby skin.
Melanistic pigment occurs on the head, nape, gut and char-
acteristically on the pectoral fins. Postanal ventral midline me-
lanophores are absent; however, pigment is added laterally with
development.
Head and preopercular spines are absent.
Malacocottus group (Fig. 230).— This group includes Malaco-
cottus and Dasycottus. Size at hatching is not known. Larvae of
this group are similar to those of the Psychwlules group with
large, blunt heads tapering to the tail. An outer bubble or layer
of skin is present in both genera and is particularly pronounced
in Malacocottus.
Pigmentation is present on the head, nape and over the entire
gut. Pigment occurs laterally on the anterior third of the tail in
Malacocottus larvae. As in the Psychrolutes group, the pectoral
fins are characteristically pigmented.
Larvae develop four preopercular spines with a fifth accessory
spine present in Malacocottus.
Hemitripterus group (Fig. 23 1 ). — This group includes the genera:
Hemiiripterus, Blepsias and Nautichthys. Hatching occurs at a
relatively large size, 7 to 1 3 mm NL. Newly-hatched larvae have
elongate, slender bodies which become deeper with develop-
ment. Nautichthys larvae have distinctively long, pigmented
pectoral fins.
Pigmentation is relatively heavy with melanophores covering
the head, dorsal surface of the gut and over the lateral body
surface except for the caudal peduncle. Nautichthys and Hem-
itripterus larvae possess distinctive pigment bands extending
onto the dorsal and ventral finfolds that are not found in other
cottid larvae.
Larvae develop four prominent preopercular spines and a
strong frontoparietal spiny ridge. This group is characterized by
delayed ossification in the larval period and a unique "honey-
comb" pattern of ossification on the head. Hemitripterus larvae
develop large bony prickles, similar to the prickle-scales found
in agonids.
Agonidae (Figs. 231 and 232). — hi least one early life history
stage of 9 of the 49 nominal species is known. Agonids hatch
at 5.5 to 8.0 mm NL. Development is a gradual transformation
to the juvenile form attained at 20 to 30 mm.
Agonid larvae are generally long and slender with relatively
long guts. Extremes of form range from short stout genera such
as Agonomalus and Bothragomis to the extremely attenuated
forms such as Ocella and Aspidophoroides. Larvae have dis-
tinctively large, fan-shaped pectoral fins.
Pigmentation varies in the family. Melanistic pigment may
be present on the head, nape, scattered over the gut and fre-
quently in bands on the postanal lateral surface of the body.
The pectoral fins are distinctively pigmented often with distal
bands of melanistic pigment. In some species (e.g. .-igonomalus,
Hypsagonus) pigmentation extends onto the dorsal and ventral
finfolds.
Larvae are characterized by spiny heads with large fronto-
parietal spiny ridges, postocular spines, and usually four large
preopercular spines. Tiny rows of spines form in small larvae
and help distinguish agonid larvae. These rows correspond to
the plates (scales) of adults.
(B.B.W.) Gulf Coast Research Laboratory, East Beach
Drive, Ocean Springs, Mississippi 39564; (W.A.L.) School
OF Natural Resources, Department of Fisheries,
Humboldt State University, Arcata, California 95521;
(H.G.M.) National Marine Fisheries Service, Southwest
Fisheries Center, Post Ofhce Box 271. La Jolla, Cal-
ifornia 92038; (W.J.R.) National Marine Fisheries Ser-
vice, Southeast Fisheries Center, 75 Virginia Beach
Drive, Miami, Florida 33149.
Cyclopteridae: Development
K. W. Able, D. F. Markle and M. P. Fahay
THE scorpaeniform family Cyclopteridae is composed of two
subfamilies (Nelson, 1976), the Cyclopterinae (lumpfishes)
with 7 nominal genera and 28 species, and the Liparidinae
(snailfishes) with 18 nominal genera and 150+ species (Table
108). Some authors have considered the subfamilies as separate
families (Gill, 1891; Garman, 1892; Jordan and Evermann.
1896-1900; Regan, 1929; Burke, 1930; Matsubara, 1955; Ueno,
1970), while others have treated them together (Boulenger, 1910;
Berg, 1940; Greenwood et al.. 1966). We follow Nelson (1976)
without prejudice; both groups appear distinct yet are clearly
sister taxa. The most compelling synapomorphy is a ventral
sucking disk (secondarily lost in some liparidines) formed from
pelvic fin rays. The cyclopterid disk differs structurally from
analogous structures in Gobiesociformes and Gobiidae (see for
example, Briggs, 1955; Ueno, 1970). Certain osteological (Ueno,
1970) and menstic differences (Table 108) between the subfam-
ilies are marked. The lumpfishes have two dorsal fins (the first
dorsal may be embedded in the skin and not externally visible
in some genera) with few total elements (4-8 spines and 8-12
rays), few anal rays (6-13) and vertebrae (23-29). The snailfishes
have a single dorsal fin with numerous elements (28-82), and
more anal rays (24-76) and vertebrae (38-86) (Table 108).
Representatives of the Liparidinae have been collected in all
oceans from the Arctic to the Antarctic. They are found from
intertidal depths to greater than 7 km (Andriashev. 1954; 1975).
However, their distribution over shallow continental shelves is
ABLE ET AL.: CYCLOPTERIDAE
429
Table 108. Nominal Cyclopterid Genera. Nlimber of Species, and Range of Meristic Characters for Each. Based primarily on data
from Burke (1930), Schmidt (1950), Ueno (1970), Andriashev (1975), Andriashev and Neelov (1976), Stein (1978), and Kjdo (1983). Dorsal fin
counts are given as dorsal spines and dorsal rays for Cyclopterinae.
No. of
species
Fin rays
Pylonc caecae
Genus
Dorsal
Anal
Pectoral
Caudal
Vertebrae
Cyclopterinae
Aptocyclus De La Pylaie
1
V, 8-
•11
6-9
19-22
9-11
15-43
27-29
Cyclopsis Popov
1
VI, 11
-12
10
23-24
10-11
9
7
Cyclopleropsis Soldatov and Popov
8
VI-VII. IC
1-12
9-13
25-28
9-11
9-10
25-26
Cyclopterus Linnaeus
1
VI-VIII, 9-
•11
9-10
19-20
11-12
36-79
28-29
Eumicrolremus Gill
14
V-VIII, 9-
•13
9-13
19-29
9-12
8-12
26-29
Letholremus Gilbert
2
VI-VII, 8-
•11
7-10
20-23
10-11
4
23-24
Pelag(xychis Lmdberg and Legeza
1
IV-V, 9-
■10
8-9
19-21
10
7
7
Liparidinae
Acanthotipans Gilbert and Burke
2
45-52
38-47
20-26
8-10
0-6
50-54
Careproclus Kroyer
47 +
40-67
32-60
17-37
6-12
0-60
47-71
Crystallias Jordan and Snyder
1
56
53
31
10
71
7
Cryslallichthys Jordan and Gilbert
2
48-53
42-44
30-35
10-12
36-40
7
Elassodiscus Gilbert and Burke
2
49-68
45-60
27-32
8-9
ca. 14-16
60
Genioliparis Andriashev and Neelov
1
53
49
19
6
7
62
Gynnichlhys Gilbert
1
—
_
25
14
—
•y
Liparis Scopoli
50-60
28-49
24-45
28-41
10-12
10-90
38-53
Lipanscus Gilbert
I
50-52
47-49
13-15
4
ca. 6
7
Nectolipans Gilbert and Burke
1
50-55
45-50
19-25
6
6-9
7
Notolipans Andriashev
3
41-57
38-53
31
10
_
50-65
Odontolipans Stein
1
51
46
17
6
7
59
Osteodiscits Stein
1
47-52
40-44
20-25
6-7
0
51-56
Paralipans Collett
27 +
48-82
42-76
14-39
3-8
5-41
57-86
Polypera Burke
3
37-44
31-34
33-37
_
200-300
7
Rhinn/ipans Gilbert
2
ca. 68
ca. 60
20-23
1-3
7-12
7
Rhodkhlhys Collett
1
56-60
54-57
16-17
10
?
ca. 65
Temnocora Burke
1
45-48
39
33-37
-
20
9
limited to the cooler waters of the arctic, antarctic and temperate
regions with the possible exception of L. ftshclsoni from the Red
Sea (Smith, 1968). The Cyclopterinae are more restricted in
their distribution, occurring exclusively in the northern hemi-
sphere's boreal and arctic waters (Ueno. 1970) where they are
usually limited to continental shelves. Although most cyclop-
teridsare benihic the cycloptenne Pelagocyclus v/7/ar/(Lindberg
and Legeza, 1955) and the liparidines Nectolipans pclagiciis and
Lipanscus nanus (Stein, 1978) are pelagic. Lipans fahricii is
considered cryopelagic in the high Arctic (Tsinovsky and
Mernikov, 1980). The cycloptenne Cyclopterus lumpus is ben-
thic during the reproductive season and pelagic at other limes
(Thorsteinsson, 1981; Able, in prep.).
Development
The available information on early life history stages is in-
adequate to allow confident generalizations about the biology
or systematics for most members of the family. This is due to
rarity of material (adults and especially larvae) and the incom-
plete understanding of cyclopterid taxonomy.
Eggs
Cyclopterid eggs are moderate to large (0.8 to 8.0 mm), de-
mersal and adhesive (Table 109). Variation in fecundity is gen-
erally related to female length (Stein, 1980a; Lisovenko and
Svetlov, 1981; Matarese and Borton, in prep.) but appears to
be a complex function of egg diameter as well (Table 109). Much
of the available information on cyclopterid eggs, summarized
in Table 109, is based on observations of ovarian eggs or oth-
erwise incomplete descriptions. It is possible, for example, that
one or more oil globules may be characteristic of all cyclopterid
eggs. Sculpturing of the chorion surface has been reported for
Lipans montagui (Mcintosh and Prince, 1890), L. tanakae
(Aoyama, 1959) and L. atlanticus (Detwyler, 1963). Pores in
the chorion have been reported for Cycloptents lumpus and L.
montagui (Mcintosh and Prince, 1890). We have found that
sculpturing of the chorion occurs in L. liparis (Fig. 234A, B, C)
and Paraliparis calidus and possibly Eumicrolremus orbis (Fig.
234D, E, ¥). Pores in the chorion are quite numerous in L.
liparis (Fig. 234B, C). Pits are present on some portions of the
egg surface of £. orbis (Fig. 234D, E, F).
Incubation is moderately long (5 to 10 weeks) in the few
reported cases (Russell, 1976; Andriashev, 1954; Matarese and
Borton, in prep.). The combination of relatively large eggs and
long incubation times results in an advanced state of develop-
ment at hatching for some members of each subfamily. In these
instances fin rays and disk are formed and notochord flexion is
underway prior to hatching (Fig. 235A, Table 110). Hatching
at an advanced state of development is characteristic for all
deep-water Liparidinae that have been relatively well studied
(Andriashev et al., 1977; Stein, 1978). Hatching may be cued
to wind induced temperature changes for some inshore Liparis
(Frank and Leggett, 1983).
Some form of parental protection, either egg hiding, paternal
guarding, or both may also be characteristic (Table 109). Some
Pacific Careproctus deposit eggs within the gill cavities of lith-
430
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 109. Summary of Egg Characteristics of Cyclopteridae.
Egg or
maximum ovar-
ian
Clutch
egg diameter
Oil
count or npe
Paternal care or
Species
(mm)
globule(s)
egg fecundity
egg deposition sites
Source(s)
Cyclopterinae
Aptocyclus ventricosus
2.3-2.4
present
3,800
paternal guarding
Kyushin, 1975; Schmidt,
1950; Kobayashi, 1962
Cyclopsis lentacularis
2.0
7
1,540
7
Lindberg and Legeza, 1955
Cyclopieropsis macalpini
5.0
?
60-70
paternal guarding,
hidden (mollusc shells)
Parr, 1926
Cyctopterus lumpus
2.2-2.7
present
15,000-200,000
paternal guarding
Zhitenev, 1970; Russell,
1976; Andriashev, 1954
Eumkrotremus birulai
3.9-4.0
present
1,230
hidden (mollusc shells)
Honma, 1956; Ueno, 1970
Eumicrotremus derjugini
4.0-5.0
7
7
7
Andriashev, 1954
Eumicrotremus orbis
2.2
present
325-477
paternal guarding
Matarese and Borton, MS
Eumicrotremus soldatori
3.1
7
4,049
•>
Ueno, 1970
Eumkrotremus spinosus
3.2-4.5
7
7
paternal guarding,
hidden (mollusc shells)
Andriashev. 1954
Lethotremus awae
1.4
7
232
7
Ueno, 1970
Liparidinae
Acantholiparis opercularis
4.8
7
1-6
7
Stein, 1980a
Careproctus sp.
5.0
7
7
hidden (lithodid crab
gill cavity)
Hunter, 1969; Vinogradov.
1950
Careproctus sp.
3.0-3.5
7
100
hidden (lithodid crab
gill cavity)
Anderson and Cailliet,
1974
Careproctus falklandwa
?
7
?
hidden (lithodid crab
gill cavity)
Balbontin et al., 1979
Careproctus longifilis
7.1
7
16
7
Stein, 1980a
Careproctus melanurus
4.2
7
534
hidden (lithodid crab
gill cavity)
Peden and Corbett, 1973
Careproctus mkrostomus
7.6
present
15
7
Stein, 1980a
Careproctus oregonensis
5.6
7
3-5
7
Stein, 1980a
Careproctus ovigerum
7.8
7
756
7
Stein, 1980a
Careproctus raslrinoides
4.5
7
7
7
Schmidt, 1950
Careproctus reinhardti
4.5
7
300
7
Collett. 1905;
Andriashev, 1954
Careproctus sinensis
5.0
7
7
hidden (lithodid
crab gill cavity)
Rass. 1950
Liparis atlanticus
0.8-1.4
present
1 ,400-3,000
paternal guarding,
hidden (algae)
Detwyler, 1963
Liparis fabricii
2.1-2.7
7
485-735
7
Andriashev, 1954
Liparis fucensis
1.0
7
7
paternal guarding,
hidden (mollusc
shell, tubeworms)
DeMartini, 1978;
Marliave. 1976
Liparis inquitinus
1.0-1.3
present
231-563
hidden (hydroids)
Able and Musick, 1976
Liparis liparis
1.4-1.7
present
7
hidden (hydroids)
Russell, 1976
Liparis montagui
1.0-1.2
present
700
hidden (red algae)
Russell. 1976;
Andriashev. 1954
Liparis pukhellus
1.5
7
941-996
?
Johnson, 1969
Liparis tanakae
1.7-1.8
present
7
hidden (sea weed)
Aoyama, 1959
Notoliparis kermadecensis
8.0
7
16
7
Neilsen, 1964
Osteodiscus cascadiae
5.3
7
1-5
7
Stein, 1980a
Paraliparis bathybius
4.5
7
422-434
?
Collett, 1905;
Andriashev, 1954
Paraliparis calidus
2.6-2.9?
7
7
?
Wenner, 1979
Paraliparis copei
2.0
7
45-86
7
Wenner, 1979
Paraliparis deani
2.0
7
0
7
Hart, 1973
Paraliparis garmani
3.5
7
190-317
oral brooding?/
paternal guarding?
Wenner, 1979; Stein.
1980a
Paraliparis gracilis
2.6-2.9
7
7
7
Marshall, 1953
Paraliparis latifrons
4.5
7
2-8'
?
Stein, 1980a
Paraliparis megalopus
4.3
7
32
7
Stein, 1980a
Paraliparis mento
2.5
7
101
7
Stein, 1980a
Paraliparis rosaceus
3.6
7
1,277
?
Stein, 1980a
Rhinoliparis barbulifer
2.5
7
7
7
Schmidt, 1950
Rhodichthys regina
3.2-4.0
7
70
7
Johnsen, 1921
ABLE ET AL.: CYCLOPTERIDAE
431
Fig. 234. Scanning electron micrographs of Lipans liparis egg (A, B, C, Zoologisch Museum Amsterdam 1 14.522. North Sea) and Eumicro-
iremus orhis egg (D, E, F) from the study by Matarcse and Borton (in prep.). The depression in A and B is the micropyle. Scale bar equals 200
n (A), 19 M (B), 4.9 M (C), 280 m (D). 28 m (E), 3.3 m (F).
odid crabs; a site which may provide both protection and water
circulation.
Larvae
In the Cyclopterinae development has only been described
for 4 of 7 nominal genera and 4 species (Table 110). Other
partial descriptions are for Aptocyclus venlricosus (Kobayashi,
1962) and Eumicrotremus spinosus (Ehrenbaum, 1905-1909;
Koefoed, 1909). In the Liparidinae, larvae of 3 of 18 nominal
genera and 10 species have been described (Table 1 10). Besides
those listed, partial descriptions have been published for Car-
eproclus georgianus (Efremenko, 1983a), Careproctus falklan-
dica and Careproctus sp. (Balbontin et al., 1979) and several
Liparis: L. atlanticus (Detwyler. 1963), L./a/>na/ (Ehrenbaum,
1905-1909; Koefoed, 1909; Johansen, 1912; Dunbar, 1947), L.
fuscensis (Marliave, 1976), L. lipans (Ehrenbaum 1904, 1905-
1909; Ehrenbaum and Strodtman, 1904; Page, 1918), L. mon-
?a.^/ (Mcintosh and Prince, 1890; Mcintosh and Mastermann,
1897; Ehrenbaum and Strodtman, 1904; Ehrenbaum, 1905-
1909; Page, 1918; Arbault and Boutin, 1968b), L. /a«aA:ae(Aoy-
ama, 1959; Kim etal., 1981), and L. /;/«/<:a/!/5( Johansen, 1912).
Morphological characters. — CycXoplend larvae typically have
flaccid skin enveloping the entire body, a short bulbous head
usually without spines, large eyes, and a trilobed lower lip. The
sucking disk forms early in development and may be present at
hatching in some forms (Fig. 235-238). The preanal length is
short and the gut is coiled. Cyclopterines may have both dorsal
fins at hatching (Fig. 235B), typically have larger disks at hatch-
ing, and usually have more pigmentation at hatching (Pig. 235-
238) than liparidine larvae. Some cyclopterine larvae develop
dermal spines that become pronounced tubercles in adults (Ueno,
1970). In many liparidine larvae the medial surface of the pec-
432
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 235. Egg (A) and larvae (B — 4.5 mm SL, C— 6.3 mm SL, ventral view) of Eumicrotremiis orbis from Matarese and Borton (in prep.) and
larvae of Cycloplerus lumpus (D— 5.0 mm SL. Damariscotta River, Maine, HML H-24029).
toral fin has numerous melanophores (Fig. 237) and during de-
velopment the fin may become bilobed (Fig. 238). The gill open-
ing decreases in size during development.
Disk size varies within each subfamily and may be related to
habitat. Pelagic forms such as the cyclopterine Pelagocychis vi-
tiazidind the liparidine L. fabricii (¥i%. 237) have small orgreatly
reduced disks. Some pelagic forms, such as Nectoliparis pclag-
icus and Lipanscus nanus lack disks entirely.
The arrangement of the cranium may offer useful insights into
cyclopterid phylogeny. Svetovidov ( 1 948) noted that the cranial
cavity extends into the interorbital space in Liparis but only
reaches the hind margin of the orbit in Cycloplerus. Our material
indicates that this character state changes ontogenetically in
Cycloplerus with the earliest stages showing the liparidine state.
Able and McAllister (1980) suggested that tooth shape ex-
hibits polarity, with trilobed teeth with equal lobes representing
the primitive condition, trilobed teeth with a larger central lobe
an intermediate condition, and simple teeth the derived con-
dition. The ontogeny of teeth in Liparis supports this statement.
All Liparis examined to date possess trilobed teeth early in
development. With continued growth the oldest teeth may be-
come simple, as in L. fabricii (Able and McAllister. 1980).
Caudal morphology and ontogeny show variation that may
prove useful for identification and phylogenetic studies. Within
liparidines caudal elements vary. For example. Stein (1978)
noted a lack of epurals in Careproclus longifilis, whose caudal
structure he considered typical of deepwater eastern Pacific li-
paridines he examined, while we note the typical presence of
Fig. 236. Larvae of Lipans (from top to bottom). Liparis allanUciis (7.9 mm NL. 47°37'N, 62°02'W, HML H-2140); ventral view of above;
L. cohem{\i.6 mm NL, Damariscotta River, Mame, HML H-24030); and an unidentified cyclopterid (5.8 mm, CALCOH 6401 Sta. 70.52).
434
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 237. Liparis fabricii larvae (from top to bottom): 16.7 mm NL, NZ 4, 74°06'N, 81°30'W, NMC 83-1 135; ventral view of above; 32. i
mm NL, NZ 292, 74°27'N, 82°03'W, NMC 83-1 136, from Arctic Canada.
two epurals in some western North Atlantic Liparis (Fig. 239)
from shallow water. Caudal development also varies. In Cy-
cloplerus (Fig. 239), Eumicrotremus (Matarese and Borton, in
prep.), and deepwater southern hemisphere liparidines (An-
driashev et al., 1977) the notochord is resorbed and flexion is
complete at hatching. In western Atlantic Liparis. especially L.
fabricii. notochord resorption and flexion are delayed as late as
50 mm SL (Fig. 237, Table 1 10).
Body proportions are also useful taxonomic characters for
larval identification. Within Liparis. larval L. fabricii are sep-
arable from other western North Atlantic Liparis by a relatively
shorter head length, smaller eye diameter, shallower body depth
and shorter preanal distance. The disk size relative to eye length
has also proven effective in distinguishing between all species
of western North Atlantic Liparis (Able et al., MS). The size of
the gill opening is difficult to measure consistently but it de-
creases as development proceeds in Liparis. suggesting that a
reduced gill opening is a derived character state.
Fig. 238. Larvae of Careproclus and Paraliparis (from top to bottom). Careproclus reinhardti. with yolk sac, 1 2.6 mm SL, Chaleur Bay, Gulf
of St. Lawrence, Canada, HML H-24031; ventral view of above; Paraliparis copei. 24.0 mm SL, St. Lawrence River estuary, Canada, HML
H-24032; and P. calidus 12.9 mm SL, St. Lawrence River estuary, Canada, HML H-24033; ventral view of above.
ABLE ET AL.: CYCLOPTERIDAE
435
tsoU
436
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 239. Caudal development of Cyclopterus tuinpus (A) 6.0 mm NL, (B) 12.5 mm SL. (C) 18.0 mm SL, HML H-3093, 43°12'N, 66°00'W;
and Liparis fahricii (D) ca. 20 mm NL, 72°30.4'N, 76°46.2'W, NMC 83-1 137; (E) ca. 34 mm NL, 74°27'N, 82°03'W, NMC 83-1 138; (F) ca. 145
mm SL, 70°07'15"N, 60°44'15"W, NMC 83-1 139). Scale bars equal one mm.
The arrangement and degree of adherence of the soft flaccid
skin of cyclopterid larvae may be of taxonomic value. In Cy-
clopterus and all western North Atlantic forms examined (Able
et al., MS) the skin conforms loosely to the entire surface of the
body. In an unidentified cyclopterid from the eastern Pacific the
skin forms a distinct bubble over the anterior portion of the
body and then adheres tightly over the posterior portion (Fig.
236).
Pi,^n7enl. — Cycloptennes are usually more heavily pigmented
at hatching (Kyushin, 1975; Matarese and Borton, in prep.; Fig.
235) than liparidines (Figs. 236, 238). An exception is L. fahricii
(Fig. 237) which has well-developed pigment. Throughout de-
velopment, all Liparis we examined from the western North
Atlantic possess melanophores on the medial surface of the
pectoral fin, on the abdomen, and a line of melanophores at the
base of the anal finfold and fin. The abdominal melanophores
ABLE ET AL.: CYCLOPTERIDAE
437
Table 1 10. Ontogeny of Character Development for Cvclopterids Based on Available Literatlire. Stage of development at hatching
indicated by pre (preflexion). and flex (flexion). X indicates event takes place before hatching.
Stage
Length (min) at development of character
al
Disk
Nostnl
Species
Source
hatching
Hatching
formation
Rexion
splitting
Poslflexion
Demersal phase
Cyclopterinae
Aptocyclus
Kyushin,
1975
flex
6.5-7.0 TL
X
X
X
7
at hatching
ventricosus
Cyclopteropsis
Parr, 1926
flex
9
X
X
?
7
7
macalpim
Cyclopterus
Fritzsche,
1978
flex
4.0-7.4 TL
X
X
?
8.0-10.0 TL?
7
lumpus
Eumicrolremus
Matarese and
flex
4.5-4.7 SL
X
X
X
55 SL
at hatching
orbis
Borton,
in prep.
Liparidinae
Careproctus
Peden and Corbett,
flex?
?
X
X
metanuius(1)
1973
Careproctus
Able et al.
, MS
flex
ca. 9.8 NL
X
X
—
17.2-21.1 SL
at hatching?
reinhardli
Liparis
Able et al.
, MS
pre
ca. 3.1 NL
3.3 NL
5.8-6.9 NL
5.4-6.3 NL
12.1-17.1 SL
7
atlanticus
Liparis
Able et al.
, MS
pre
ca. 3.4 NL
3.7-6.0 NL
8.1-8.5 NL
5.
1-9.4 NL
14.9-19.0 SL
14 TL
inquilinus
Liparis coheni
Able et al.
, MS
pre
ca. 5 NL
ca. 5.0 NL
8.5-9.3 NL
7.8-9.6 NL
19.9-20.7 SL
29-36 TL
Liparis gihbi/s
Able et al.
, MS
pre
4.8 NL
7.4 NL
7.4-10.3 NL
12.
7-15.4 NL
20.0-41.7 SL
9
Liparis
Able et al.
, MS
pre
ca. 8 NL
8.6-11.5 NL
11.9-13.4 NL
14.
1-17.2 NL
48.2-52.1 SL
7
fabncii
Liparis
Able et al.
, MS
pre
7
9
?
7
20.8-27.0 SL
7
tunicatus
Paraliparis
Able et al.
, MS
flex?
?
—
X?
18.6-20.7 SL
at hatching?
calidus
Paraliparis
Able et al.
, MS
flex?
?
—
X''
—
ca. 18.0 SL
at hatching?
copei
are variable, with some species lacking meianophores on the
ventral surface behind the disk (Fig. 236) while in L. fabricii
they are prominent (Fig. 237). A second row of meianophores
occurs on the edge of the anal finfold in preflexion L. fucensis
(Marliave, 1976). The early appearance of meianophores on the
lateral surface of the tail pnor to beginning of notochord flexion
is diagnostic for L. atlanticus (Fig. 236) among western North
Atlantic Liparis with the exception of L. fabricii. Liparis fabricii.
unlike other Liparis examined, has numerous, stellate meia-
nophores over most of the body and these become increasingly
numerous with development (Fig. 237). By late flexion the pe-
ritoneum is completely black, the pectoral fins and head are
very dark, and oblique patches of meianophores are apparent
on the dorsal and anal fins (Fig. 237). All of these patterns are
unique to this species and suggest that it may be relatively iso-
lated within the genus. Generally, pigmentation patterns should
be used with caution since geographical vanation does occur,
as for L. gibbus (Able et al., MS).
Ontogenetic schedule.— On the basis of current information, it
appears that certain developmental landmarks are useful for
distinguishing between groups of cycloptends and may, in some
instances, reflect relationships. The degree of development at
hatching is variable both between and within subfamilies (Table
110). All cyclopterines studied hatch late in development, at
relatively large sizes, when many developmental characters are
nearly complete (see Fig. 235). Embryonic development is more
variable within the liparidines (Table 1 10); some Careproctus,
Paraliparis {Fig,. 238) and other deepwater forms from the south-
em hemisphere (Marshall, 1953; Andriashev et al., 1977) ap-
parently hatch late in development, at large sizes while shallow
water Liparis studied to date hatch as preflexion larvae (Able
et al., MS).
Within Liparis, the development of several characters occurs
over a wide size range (Table 1 10). For example, in L. atlanticus
and L. inquilinus hatching, disk formation, nostril splitting, flex-
ion and postflexion and assumption of demersal habitat occur
at relatively small sizes, while in L. fabricii all of these events
are delayed until larger sizes. Other species (L. coheni, L. gibbus)
are intermediate. Liparis fabricii. which shows the most delayed
development, may remain pelagic throughout its life (Able and
McAllister, 1980; Tsinovsky and Mel'nikov, 1980). While some
of this variation may be explained by the variation in egg size
it can not account for the great differences observed. We suggest
that delayed development is associated with delayed assumption
of the demersal habitat and that this represents neoteny.
(K.W.A.) Biological Sciences and Center for Coastal and
Environmental Studies, Rutgers University, New
Brunswick, New Jersey 08903; (D.F.M.) Huntsman
Marine Laboratory, St. Andrews, New Brunswick EOG
2X0, Canada; (M.P.F.) National Marine Fisheries Ser-
vice, Northeast Fisheries Center, Sandy Hook
Laboratory, Highlands, New Jersey 07732.
Scorpaeniformes: Relationships
B. B. Washington, W. N. Eschmeyer and K. M. Howe
THE order Scorpaeniformes is a large, morphologically di-
verse group containing about 20 families (depending on
classification used), 250 genera, and over 1,000 species. The
order is defined by the presence of a suborbital stay, a posterior
extension of the third infraorbital bone which in nearly all species
is firmly attached to the preopercle. Infraorbital bones for many
scorpaeniform groups were discussed most recently by Poss
(1975). Many workers have suggested that the stay may have
evolved independently (Matsubara, 1943;Quast, 1965; Green-
wood et al., 1966; Poss, 1975; and Nelson, 1976).
Relationships
The higher classification of the Scorpaeniformes remains con-
troversial and uncertain, both in terms of monophyly and in
the definition of families and their relationships. Confusion ex-
ists not only at the subordinal levels, but also at lower taxonomic
levels. For example, between 1 and 17 families of cottids have
been recognized by previous workers.
Two workers presented hypotheses of relationships within the
Scorpaeniformes. Matsubara (1943), in a detailed study of Jap-
anese scorpaenoids based on osteological and anatomical char-
acters, briefly treated relationships of scorpaenoids to other scor-
paeniforms. His graphic presentation of relationships is shown
in Figure 240. Several lineages are recognizable: 1) the Hexa-
grammidae, Anoplopomatidae, and "generalized" scorpaenids;
2) Peristediidae, Triglidae, and Dactylopteridae; 3) "special-
ized" scorpaenids, Bembridae, Platycephalidae, and Hoplich-
thyidae; 4) Cottidae and Agonidae; and, 5) Cyclopteridae and
Liparididae. In 1955, Matsubara refined his hypothesis of re-
lationships and presented a classification with categories equiv-
alent to three suborders, several superfamiliesand included fam-
ilies as follows:
Cottida
Cottina
Scorpaenicae
Scorpaenidae, Synanceiidae, Congiopodidae
Hexagrammicae
Anoplopomatidae, Hexagrammidae
Plalycephalicae
Parabembridae, Bembridae, Platycephalidae, Hop-
lichthyidae
Cotticae
Cottidae, Psychrolutidae
Agonicae
Agonidae, Aspidophoridae
Triglicae
Triglidae, Peristediidae
Dactylopterina
Cephalacanthidae
Cyclopteridae
Cyclopteridae, Liparididae
Quast (1965) presented a notably different hypothesis of re-
lationships of the scorpaeniforms. His work was based on char-
acters which were useful in comparisons with the hexagrammids
and included many characters taken from the earlier works of
Gill (1888), Regan (1913a) and Berg (1940). Quast proposed that
the Scorpaeniformes included three basic lineages: 1 ) the cottid-
hexagrammid (including the Cyclopteridae and Agonidae); 2)
the anoplopomatid; and, 3) the scorpaenoid (including all other
families). Quast (1965) did not incorporate his recommended
revisions in his formal synopsis of scorpaeniforms because he
believed that the cottoids and anoplopomatids were still in need
of intensive study.
Several studies of particular character complexes have also
contributed to understanding of relationships within the order.
Freihofer (1963), in a study of patterns of the ramus lateralis
accessorius and associated nerves in teleosts, found three pat-
terns of nerves in scorpaeniforms which suggested three groups:
1) the Scorpaenidae and Synanceidae; 2) the Hexagrammidae,
Cottidae, and Liparididae; and, 3) the Anoplopomatidae. These
groupings seem to support Quast's hypothesis of relationships
but many families were not examined by Freihofer. Hallacher
(1974) provided a summary of gasbladder muscles in the scor-
pionfish genus Sebastes and included observations on other
scorpaeniforms. Matsubara (1943) treated this feature for Jap-
anese scorpaenoids. Hallacher recognized four states of the ex-
trinsic muscle in scorpaeniforms. His characters were based on
the connections, or lack of connections, of this muscle between
the cranium, pectoral girdle, vertebral column, and the gas-
bladder. His observations partially supported Matsubara's hy-
pothesis of scorpaeniform lineages.
Scorpaeniform fishes have been considered as pre-perciforms
or as perciform derivatives but their relationship to other fishes
remains uncertain. Many workers have argued that the Scor-
paeniformes evolved from a "generalized" perciform ancestor
because of striking similarities in general body form, and ana-
tomical and osteological characters of generalized scorpaenids
and perciforms (Gill, 1888; Regan, 1913a; Taranets, 1941; Mat-
subara, 1943; Gregory, 1959; Quast, 1965; Gosline, 1971; Lau-
der and Liem, 1983). Others (Greenwood et al., 1966; Nelson,
1976) have tentatively placed the Scorpaeniformes as a distinct
pre-perciform group of the Acanthopterygians.
As previously mentioned, several authors have suggested that
the Scorpaeniformes may be polyphyletic and hence, derived
from several groups. Greenwood et al. (1966) noted that some
scorpaeniforms share similarities of the parietals and cheek
muscles with cods, while others share similarities with toad-
fishes, and still others with perciformes. Freihofer (1970), on
the basis of nerve evidence, suggested that gobiesocids were
related to cottoids, especially liparidids. Although several au-
thors have suggested that the suborbital stay may have evolved
more than once in the Scorpaeniformes, little consideration has
been given to the hypothesis that other groups of fishes may
have lost the suborbital stay. Within the Scorpaeniformes, sev-
eral groups show a reduction or loss of the suborbital stay.
Groups which have lost the circumorbital bones, and possibly
a suborbital stay (e.g. gobiesocids, callionymids, lophiiforms
438
WASHINGTON ET AL.: SCORPAENIFORMES
439
and gobioids) should not be excluded from consideration of
relationships to some scorpaeniform groups.
In summary, the limits of the order, suborders, families and
distribution of families in the suborders are the subject of con-
siderable disagreement among current workers. These problems
will not be resolved without a worldwide revision of the order.
At this point, we assume that the order is monophyletic. For
the purposes of summarizing information on this order, we treat
two broad suborders: the Scorpaenoidei and the Cottoidei. We
consider these groups as a convenient way to discuss disagree-
ments in classification of specific groups and hypotheses of re-
lationships; we do not propose that they are monophyletic groups.
Suborder Scorpaenoidei
For this paper, we recognize the Scorpaenoidei to include the
following families: Scorpaenidae (broad sense of Matsubara,
1943), Triglidae, Peristediidae, Bembridae, Platycephalidae,
Hoplichthyidae, and Dactylopteridae. Some of these families
have been assigned to separate suborders or superfamilies and
the dactylopterids have often been placed in a separate order
(Quasi, 1965; Nelson, 1976; Lauder and Liem, 1983).
Meristic features and approximate number of species for in-
cluded groups are provided in Table 111. Data have been drawn
from many sources and may not be complete for some genera
or may omit extremes found in abnormal individuals.
Matsubara's work (1943) is the most thorough study of scor-
paenoids to date. His hypothesis of relationships (Figure 240)
is based on a wide variety of characters including those of the
infraorbital bones, suspensorium, hyoid apparatus, cranium,
pectoral girdle and gasbladder. Matsubara included 14 subfam-
ilies in his family Scorpaenidae. He recognized three large ge-
neric groups or lineages within the scorpaenoids which he la-
beled: Sebasles-slem, Scorpaena-siem and Cocotropus-stem. His
Sehasles-stem contains two subfamilies, the Sebastinae and
Neosebastinae which were viewed as the most primitive or "gen-
eralized" of the scorpaenoids. His second group, the Scorpaena-
stem, includes five subfamilies: Scorpaeninae, Pteroinae, Setar-
chinae, Sebaslolobinae, and Plectrogeninae. The third group,
the Cocotropus-ilem, includes six subfamilies: Apistinae, Con-
giopinae, Aploactinae, Minoinae. Pelorinae, and Erosinae. The
latter two groups were considered "specialized" or derived rel-
ative to the Sebastes-%\em. Other worJcers (Greenwood et al.,
1966; Nelson, 1976; Poss and Eschmeyer, 1978) have departed
from Matsubara's classification of the Scorpaenidae by elevating
some subfamilies of Matsubara to family status. In addition,
the Congiopodidae [but not Matsubara's Congiopinae (sic)] has
been recognized as a separate family in a monotypic suborder
by Greenwood et al. (1966) and Nelson (1976) and as a super-
family by Quast (1965). Other scorpaenoid groups not treated
by Matsubara (1943) have been given separate status within the
Scorpaenoidei by the aforementioned workers, and include the
Caracanthidae and Pataecidae. In his later work on fish hier-
archy, Matsubara (1955) recognized three families of scorpae-
nids which basically correspond to his earlier three "stem" groups.
We basically follow the phylogenetic hypotheses of Matsubara
(1943, 1955) in presenting general trends in relationships within
the suborder. The following discussion highlights groups where
problems or disagreements about relationships are persistent.
A phylogenetic approach based on information presented here
would result in family and subfamily lines being interpreted
quite differently. However, we believe presentation of a new
Fig. 240. Schematic representation of scorpaeniform relationships
from Matsubara ( 1 943).
classification is premature; a thorough study of the scorpaenoids
is required on a worldwide basis.
The Sebastinae is currently considered to be the most prim-
itive or generalized group of scorpaenoids because of the in-
complete suborbital stay in Sebastes, weak head spination, and
general body plan similar to the percoids (Matsubara, 1943 and
others). Matsubara (1943) proposed that Sebastes was the most
generalized genus within the subfamily with a transition series
to Helicolenus. Eschmeyer and Hureau (1971) and Barsukov
(1973) believed that Matsubara's transition series is reversed
with Helicolenus the most generalized genus and Sebastes being
a relatively derived form.
The subfamily Scorpaeninae with its 1 50 genera is considered
a "catch-basket" subfamily, and there is no certainty that it is
monophyletic.
Matsubara (1943) noted that the Setarchinae lack a basi-
sphenoid as do cottoids and that the second and third actinosts
intervened between the hypercoracoid and hypocoracoid. He
concluded from these observations that the Setarchinae and
cottoids shared a common ancestor. However, Eschmeyer and
Collette (1966) disagree. In their review of the Setarchinae, a
small basisphenoid, connected only by cartilage, was found in
cleared and stained specimens; they stated that Matsubara's
conclusion was untenable.
Matsubara (1943) suggested that the genus Sehastolobus was
closely related to the genus Plectrogenium (subfamily Plectro-
geninae) because of their shared lack of gasbladders, notched
pectoral fins and prominent rows of spines along the sides of
their head. He further noted (1943:160): thai "Plectrogenium
440
ONTOGENfY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 111. Meristic Features for Suborder Scorpaenoide
. Parentheses show
rarer counts;
abnormal specimens not i
ncluded.
Dorsal fin rays
a
nal fin rays
[Genera
Pectoral
Pelvic
Taxon
Spe
les
Spines
Sofl rays
Tolal
Spines
Soft rays
rays
rays
Vertebrae
Scorpaenidae
Sebastinae
Helicolenus
12
(11) 12(13)
11-14
23-26
3
5-6
(17) 18-20
1 + 5
25
Hozukius
2
12
11-12
23-24
3
6
18
I + 5
(25)26
Sehastiscus
4
(11) 12(13)
11-13
22-25
3
(4) 5 (6)
17-19
1 + 5
25(26)
Sebastes
110
(12) 13-16
11-17
24-30
3
5-10
15-21
I + 5
(25)26-31
Scorpaeninae
[15] 150
12-13
8-10
21-23
3
5
14-21
I + 5
24
Sebastolobinae
Adelosebastes
1
13
13
26
3
5
21
I + 5
26
Trachyscorpia
2
12-13
8-9
20-22
3
5
20-24
I + 5
25-26
Sebastolobus
2
15-17
8-10
23-27
3
(4) 5 (6)
20-24
I + 5
27-30
Plectrogeninae
[1]
2
12
(6)7
(18) 19
3
5
20-24
I + 5
26
Pteroinae
[5]
17
12-13
9-12
22-24
2-3
5-9
13-21
I + 5
24
Setarchinae
[3]
5
(11) 12(13)
9-10(11)
21-23
2-3
5-6 (7)
19-25
I + 5
24
Neosebastinae
[2]
12
13
(6) 7-8 (9)
19-22
3
(5)6
19-23
I + 5
25-26
Apistinae
Apistus
1
-2
14-16
8-10
23-25
3
6-8
10-12 + 1
I + 5
(25)26
Cheroscorpaena
1
13
7-9
20-22
3
6-7
9 + 3
I + 5
26
Minoinae
[1]
10
8-11 (12)
(8)9-14
19-24
2
7-11
11 + 1
I + 5
(24) 25-27
Choridactylinae
Chohdactylus
2
12-15
8-9(10)
21-23(24)
2
8-10
9 + 3
I + 5
26-28
Inimicus
8
15-18
(5) 6-9
23-26 (27)
2
8-13
10 + 2
I + 5
27-30
Synanceiinae
Synanceia
4
-5
(12) 13-14(15)
4-7
18-20
3
4 (5*)-6 (7)*
11-19
1 + 4-5
24
Erosa
1
14
5-6
19-20
3
5-6
14-16
1 + 4
(24) 25 (26)
Dampierosa
1
12(?13)
(8)9
21
2
6
12
I + 4
—
Pseudosyna nceia
1
(15) 16-17
4*-6*
19*-21*
3
7*-8*
14-15(16)
I + 3
26-27
Leptosynanceia
1
16
5*
21
3-4
5*-6*
13-15
I + 4
28
Trachwephatus
1
(11) 12(13)
12*- 14*
24»-26*
2
12*-14*
14-15
I + 5
(28)29(30)
Tetrarogidae'
Ablabys
3
15-17
7-11
23-27
3
5-9
11-13
I + 5
26-28
Centropogon
3
15-16
7-9
23-24
(2)3
5(6)
13-15
I + 5
26-27
Coccolropsis
1
14-16
5-6
19-22
3
(3)4(5)
11
I + 3
25-27
Cottapistus
1
13-15
5-7
18-21
3
5-6
(13) 14(15)
I + 4
(24)25
Gtyptauchen
1
16-18
6-7
23-25
3
5
13-15
I + 5
26-28
Gymnapisles
1
13-14
7-9
20-22
3
5-6
11-12
I + 5
27-29
Liocramum
1
13-14
6-9
20-22
3
5-6
13-15
I + 4
24-25
Neocentropogon
4
13-15
7-8
20-22
3
5-7
13-16
I + 5
25
Notestes
1
14-16
8-10
22-25
3
5
11-14
I + 5
27-28
Ocosia
6
14-17
7-9
23-24
3
5
12-13
I + 5
26-30
Paracentropogon
2
-4
13-15
6-8
20-22
3
(3)4(5)
9-12
I + 4
25-27
Richardsonichthys
1
12-13
5-8
18-20
3
(5) 6-7
14-16
I + 5
24-25
Sny derma
2
12-14
10-11
22-24
3
5-6
13-15
I + 5
24-28
Tetraroge
2
13(14)
6-9
20-22
3
(4)5(6)
11-12
1 + 5
24-26
Vespicula
ca
5
3 + 8-13
3-8
18-21
3
3-5
10-14
I + 5
24-26
Paetaecidae'
Aelapciis
3
19-22
11-13
30-34
7-9
4-5
8
0
35-37
Neopalaecus
1
19-23
7-10
29-31
5-7
3-4
8
0
34-37
Pataecus
1
22-25
14-17
38-40
9-11
4-7
8
0
41-44
Ganthanacanthidae'
Gnathanacanthus
1
11-13
9-11
20-23
3
8-9
10-12
I + 5
28-30
Congiopodidae'
Alerlichthys
1
14-16
10-13
26-27
2
7-9
9
1 + 5
30-31
Congiopodus
4?
16-20
11-14
28-33
0-2
7-10
(8)9
I + 5
36-38
Perryena
1
14-15
8-9
23-24
3
5-6
12
I + 5
28
Zanchlorynchus
1
7-9
12-14
19-22
0
8-10
8-9
I + 5
36
Aploactinidae-
Acanthosphex
1 +
11-13
7-11
19-22
1(2)
6-8
9-10
I + 2
24-26
Adventor
1
3+10
7-9
21-23
1
8-10
12-14
1 + 2
27 (28)
Aploactis
1
12-15
11-15
24-28
1-3
10-12
11-14
I + 2
28-30
Aploactisoma
1
13-15
12-16
26-29
1
9-13
10-11
I + 2
30-33
Bathyaploactis
2
14-15
7-9
21-23
3-4
5-9
10-12
I + 2
25-28
Cocolropus
10
12-15
7-12
19-24
1-2
6-9
11-14
I + 3
25-28
WASHINGTON ET AL.: SCORPAENIFORMES
441
Table 111. Continued.
[Genera]
Spec
les
Dorsal tin rays
Anal fin rays
Pecioral
rays
Pelvic
rays
Taxon
Spines
Soft rays
Total
Spines
Soft rays
Vertebrae
Ensphe.x
3 +
10-12(13)
9-16
21-28
1(2)
9-15
11-15
1 + (1)2
27-31
Eschmeyer***
1
8
13
21
3
8
19-20
I + 3
24
Kanekonia
3
11-13
7-10
20-22
1(2)
7-9
13-16
I + (1)2
25-26
Neaploactis
1
4 + 7+1
9-10
21-22
1(2)
7-9
12
I + 3
26
Paraploaclis
7
12-15
8-11
22-24
1(2)
7-10
13-15
I + 3
26-28
Perislrominous
1
12-13
10-11
22-24
0-2
7-10
14-15
I + 3
26-27
Prosoproaus
1
12
8
20
2
7
13
I + 3
26
Ptarmus
2
13-16
7-10
21-23
2
4-7
9-10
I + 2
25-30
Sihenopus
1
3 + 9
8-10
20-22
1
7-9
14-15
I + 2
26
Xenaploaclis
3
3 + 9-10
8-9
21-22
1
9-10
14-15
I + 3
27-28
Caracanthidae
[1]
4
6-8
11-14
18-23
2
11
12-15
I + 2-3
24
Triglidae
[ca. 10] 80
7-11
10-19
18-26
0-1
11-18
11-16 + 3
I + 5
34-38
Penstediidae
[3] 40
7-9
16-23
24-31
0
16-23
+ 2
I + 5
Bembridae
Parabembras
1
9; 11-12
8-9
18; 20
3
5
21
I + 5
26
Bembradon
1
6
14
20
0
14*-15*
23
I + 5
Bembras
1
10-11
12
21
0
14*-15*
I + 5
Bembradium
1
8-9
12
20-21
0
10*-11*
24-27
I + 5
26
Platycephalidae
[18] 60
6-9(10)
11-15
18-23
O-I
10-14
16-22
I + 5
27
Hoplichthyidae
[1]
10
5-6
14-16
19-21
0
16-18
13-14 + 3-4
I + 3-5
26
Dactylopteridae
Dactylopterus
1
7
8
15
0
6*
34-37
I + 4
22
Dactyloptena
6
7*-8**
8(9)
15(16)
0
6(7)*
28-35
I + 4
22
• Last ray single (usual condition is a double ray).
" I + 0 + V + I or I + I + V + I = 7-8 spines.
**" Placement uncertain.
' Data supplied by Poss.
- From Poss (1982).
nanuni is closely related to the bembrids," and that "it is very
probable, therefore, that the platycephalids, bembrids and hop-
lichthyids arose from an ancestor not very unlike the scorpaenid.
Plntrogeniuin naiiuin." Matsubara and Ochiai (1955) present
additional characters which support this view. Other characters
observed by one of us (WNE) which support this conclusion
include, "similar caudal skeletons and scales" [comparison of
bembrid Parabembras ciirtis (SU 49456, cleared and stained)
and Pleclrogeiiiu?}!]. At present, this available evidence suggests
that the scorpaenid subfamilies Sebastolobinae and Plectrogen-
inae and the families Bembridae, Platycephalidae and Hoplich-
thyidae may form a monophyletic assemblage.
Another scorpaenid subfamily, the Apistinae, also has ques-
tionable relationships within the family. Matsubara ( 1 943) placed
Apislus at the base of his Cocolropiis stem which led to a number
of specialized scorpaenid groups. However, the Apistinae have
a "unique", bilobed gasbladder with an intrinsic muscle, unlike
other scorpaenids (Matsubara, 1943; Hallacher, 1974). The trig-
lids and peristediids possess a similar gasbladder. Other char-
acters which appear to unite the Apistinae, Triglidae, and Per-
istediidae include elongate pectoral fin rays, 1 to 3 lower free
pectoral fin rays(l in Apistus. 3 in Chcroscorpaena. 3 in triglids,
and 2 in peristediids) and shape and expansion of the head bones
(especially the infraorbidal bones). These characters suggest that
the scaled, less bony-headed Apistinae may be the primitive
sister group of the Triglidae and Penstediidae. This would in-
volve the independent development of a moveable, preorbital
bone with long spine in the Apistinae. If in fact, the Apistinae
forms part of a monophyletic assemblage with the triglids and
peristediids, a change in classification would be warranted.
Matsubara (1943) recognized five Japanese genera within his
subfamily Congiopinae (sic) which more recently have been
placed in the tentative scorpaenid subfamily, Tetraroginae (Poss
and Eschmeyer. 1975; see also Smith, 1958b). The presently
recognized Congiopodidae is considered to contain 7 to 9 species,
all of which are confined to the Southern Hemisphere (More-
land, 1960; Hureau, 1971). Moreland (1960:241) slated: "the
Congiopodidae show relationship with the Scorpaenidae, par-
ticularly with Snyderina and Ocosia from Japan [studied by
Matsubara (1943)], and are clearly derived from a scorpaenid
stock of perhaps Indo-Pacific origin."
We tentatively include the Dactylopteridae in our discussion
of the Scorpaenoidei, however relationships of these fishes re-
main uncertain. They have been variously placed in their own
order (Regan, 1913a; Berg, 1940; Greenwood et al., 1966; Lau-
der and Liem, 1983) and as a suborder of the Scorpaeniformes
(Gill, 1888; Nelson, 1976). Many workers have noted that the
dactylopterids differ markedly from scorpaeniforms in a number
of osteological characters such as: 1) nasals fused into a median
plate; 2) very large extrascapulars; and, 3) mesethmoid and in-
tercalar absent.
Matsubara (1943) suggested that despite these notable differ-
ences, the dactylopterids possess the characteristic suborbital
arrangement of bones of the generalized scorpaenids and triglids.
.•Accordingly, Matsubara (1943) placed them near the triglids
and peristediids, evolving from a generalized scorpaenid ances-
tor. One of us (WNE) has noted that the gasbladders of the
triglids and dactylopterids appear to be similar, with anterior
and posterior lobes and very large intrinsic muscles (sec Evans,
1973, for information on triglids). In dactylopterids the gas-
442
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 12. Meristic Features for Suborder Cottoidei. Dashes [— ] indicate data not available.
Dorsal fin
A
lal tin
No. of
species
Pectoral
fin
Pelvic
fin
Vertebrae
Genera
Spines
Rays
Spines
Rays
References
Agonidae
Agonomalus
4
8-10
5-8
0
11-12
11-12
1,2
—
Howe'
Agonopsis
3
6-11
6-9
0
7-12
12-15
1,2
39-42
Howe'
Agonus
7
5-11
5-14
0
5-17
13-19
1,2
36
Howe'
Anoplagonus
2
0
4-6
0
4-5
8-12
1,2
41-45
Howe'
Aspidophoroides
3
0
4-7
0
4-7
9-16
1,2
38-40
Howe'
Balhyagonus
4
5-8
5-8
0
5-9
14-16
1,2
40-46
Howe'
Bolhragonus
2
2-5
4-6
0
4-6
10-12
1,2
31-38
Howe'
Brachyopsis
2
7-9
7-9
0
10-14
14-15
1,2
—
Freeman, 1951
Hypsagonus
2
7-11
5-7
0
9-11
12-14
1,2
36
Howe'
Ocella
6
7-13
6-9
0
7-18
14-18
1,2
33-39
Howe'
Odontopyxis
1
3-6
5-7
0
5-7
12-15
1,2
37-42
Howe'
Pallasina
3
5-9
6-9
0
9-14
10-13
1,2
45-47
Howe'
Percis
2
5-7
5-9
0
7-9
11-12
1,2
42
Howe'
Sarrilor
2
6-9
5-8
0
6-8
13-17
1,2
—
Howe'
Stellenna
1
6-8
5-8
0
7-9
16-19
1,2
34-37
Howe'
Tilesina
2
19-21
7-10
0
25-28
15-16
1,2
—
Freeman, 1951
Xeneretmus
4
5-9
6-9
0
5-9
12-17
1,2
39-43
Howe'
Anoplopomalidae
.-1 noplopoma
1
17-30
16-21
III
15-19
I. 5
61-66
Richardson and Washington, 1980;
Andriashev, 1954; Miller and Lea,
1972; Howe'
Erilepis
1
12-14
16-20
III
11-14
16-19
1,5
45-46
Andriashev. 1955a
Comephoindae
Comephorus
2
6-9
28-34
0
27-36
10-15
-
48-50
Taliev, 1955
Cottidae
Alcichthys
1
9-10
14-17
0
13-16
15-16
1,2-3
33-36
Watanabe. 1960
Antipodocottus
2
8
14-15
0
11-12
18-19
1,2
—
Nelson, 1975
*Archaulus
1
9-10
28-29
0
22-23
16
1,3
—
Howe'
Archistes
1
10
23
0
18
15-16
1,3
—
Jordan and Gilbert, 1899
Argyrocollus
1
8-9
14-19
0
11-16
13-14
1.2-3
35-36
Watanabe, 1960
*Anedieltichthys
1
7-9
12-13
0
9-11
21-23
1,3
—
Howe'
*Artedieltina
1
_
—
—
—
—
—
—
Artedielliscus
1
—
—
—
—
—
—
—
Artediellus
15
6-9
11-14
0
10-14
20-24
1,3
28-30
Howe'; Leim and Scott, 1966
Arledius
7
7-10
12-18
0
9-14
13-17
1.2-3
30-35
Howe and Richardson. 1978; Wash-
ington. 1981
Ascelichthys
1
7-10
17-19
0
13-16
16-18
0
33-36
Howe and Richardson. 1978
Asemichthys
1
9-11
14-16
0
15-16
16-18
1,3
33-35
Howe and Richardson. 1978
Astrocollus
2
7-10
12-14
0
10-12
15-17
1,2-4
28-29
Watanabe. 1960. 1976
Bero
1
9-10
15-16
0
13-15
15-16
1,2
32-35
Watanabe. 1960
Blepsias
2
6-10
20-26
0
18-22
11-17
1,3
37-39
Howe and Richardson, 1978;
Watanabe, 1960
Chitonolus
1
8-11
14-17
0
14-17
16-18
1,2-3
35-36
Howe and Richardson, 1978
Ctinocottus
5
7-10
13-17
0
9-14
12-15
1,3
31-35
Howe and Richardson, 1978; Wash-
ington, 1981
Cottiusculus
3
7-10
11-15
0
9-15
19-22
1,3
24-29
Watanabe, 1960
Coitus
35±
4-10
14-23
0
10-18
10-19
1,2-5
31-39
Howe-
*Crossias
1
8-10
17-20
0
10-16
14-16
1,3
—
Soldatovand Lindberg. 1930;
Watanabe. 1960
*Damma
1
8-10
13
0
12
16
1,2
_
Watanabe, 1960
*Enophrys
6
7-9
9-14
0
6-13
15-19
1. 3-4
29-33
Sandercock and Wilimovsky, 1968;
Howe and Richardson, 1978
Furcina
2
8-11
15-20
0
13-18
13-15
1,2
32-37
Watanabe, 1960
Gymnocanthus
6
9-12
13-18
0
14-20
15-21
1,3
33-40
Howe and Richardson, 1978; Leim
and Scott. 1966; Watanabe. 1960;
Wilson. 1973
HemUepidotiis
6
8-11
18-20
0
13-16
14-17
1,4
35-37
Howe and Richardson, 1978
Hemitripierus
3
11-19
11-14
0
12-15
18-22
1.3
37-41
Howe and Richardson, 1978;
Leim and Scott, 1966
Icelinus
9
8-12
12-18
0
10-17
15-19
1,2
33-39
Howe and Richardson, 1978; Yabe
etal. 1980; Peden. 1981
*Icelm
13
7-10
17-24
0
13-20
17-20
1,3
37-44
Howe'
Jordania
1
17-18
15-18
0
22-24
13-15
1,4-5
46-48
Howe and Richardson, 1978
WASHINGTON ET AL.: SCORPAENIFORMES
443
Table 112. Continued.
Dorsal fin
Anal fin
No. or
species
Pectoral
fin
Pelvic
fin
Vertebrae
Genera
Spines
Rays
Spines
Rays
References
Leiocottus
1
9-10
16-17
0
15-20
18
1,3
35-36
Howe and Richardson, 1978
Leptocollus
1
6-8
15-20
0
15-20
17-20
1.4
35-39
Howe and Richardson, 1978
Megalocottus
2
8-10
12-15
0
11-13
16-18
1,3
—
Howe'; Soldatov and Lindberg.
1930
Mesocotlus
1
8-9
14-15
0
10-12
—
1,3-4
—
Soldatov and Lindberg, 1930
Microcottus
1
7-9
12-14
0
10-12
14-17
1,3
32-34
Howe'
Myoxocephalus
18
8-12
10-20
0
8-16
14-19
1,3
34-46
Howe-; Andriashev. 1954
Naulichlhys
3
7-10
19-30
0
14-21
13-17
1.3
35-41
Peden, 1970
Ocynectes
2
9-13
12-17
0
6-11
13-15
1,2
29-31
Watanabe, 1960
Oligocollus
4
7-10
15-20
0
9-15
12-15
1,3
33-37
Washington, 1981; Howe and Rich-
ardson, 1978
Orthonopias
1
8-9
15-18
0
12-15
13-15
1,3
33-35
Howe and Richardson, 1978
Pancelinus
1
12-13
19-20
0
23-24
14-15
1,5
42
Howe and Richardson, 1978
Phallocollus
1
10-12
22-24
0
22-25
14-16
1,3
—
Howe'
Porocotlus
6
8-10
13-18
0
11-18
13-19
1.3
34-38
Howe'; Andriashev. 1954;
Watanabe. 1960
Pseudoblennius
6
8-11
15-21
0
12-18
13-16
1,2
32-38
Watanabe. 1960
Radulinopsis
7
9-10
14-15
0
14-15
16-17
1,3
—
Soldatov and Lindberg. 1930
Radulinus
5
8-11
20-23
0
21-25
17-20
1,3
38-40
Howe and Richardson, 1978
Rhamphocotlus
1
7-9
12-14
0
6-8
14-16
1.3-4
26-28
Howe and Richardson. 1978
Ricuzenius
2
8-11
14-20
0
10-19
15-19
1.2-3
28-32
Watanabe. I960, 1976; Jordan
andStarks. 1904
Scorpaenichthys
1
8-12
15-19
0
11-14
14-16
1,4-5
35-37
Howe and Richardson, 1978
Sigmisies
2
8-10
19-26
0
14-20
13-15
1,3
34-36
Howe and Richardson, 1978
Stelgislrum
T
8-9
17-19
0
12-14
14-16
1.3
36
Howe'
Slernias
1
10-11
22-24
0
22-24
16-18
1.3
44-46
Howe and Richardson, 1978
Sllegicottus
1
9
19
0
17
18
1.3
—
Howe and Richardson, 1978
Sllengis
3
7-11
13-16
0
11-15
11-20
1,2
29-35
Watanabe. 1960
Synchirus
I
8-10
19-21
0
18-21
21-24
1,3
38-39
Howe and Richardson. 1978
*Taurocollus
->
11
15-16
0
12-13
19
1,3
_
Howe'; Taranetz. 1935
Thecopterus
1
10
14
0
11
20
1,2
—
Howe'
Thyriscus
1
10
21
0
17
15
1,3
38-39
Howe'
Trachydermus
1
8
18-19
0
16-17
16-17
1,4
34-36
Watanabe. 1960
Tnglops
9
9-13
20-32
0
19-32
15-24
1,3
44-54
Howe'; Andriashev. 1954;
Watanabe, 1960; Leim and
Scott, 1966
Vellilor
1
10
18-20
0
17-20
13-15
1,2
36-39
Watanabe, 1960
Zeslicelus
1-2
5-7
10-13
0
8-11
19-21
1,2-3
25-26
Howe and Richardson, 1978
Cottocomephoridae
Asprocoltus
5
5-8
12-17
0
11-16
13-17
1,3-4
30-34
Taliev, 1955
Ahyssocottus
5
3-7
10-16
0
8-15
12-18
1,2-3
31-34
Taliev, 1955
Bat rachoco! Ills
4
5-8
14-19
0
10-15
14-19
1,3
32-37
Taliev, 1955
Conmclla
-)
5-7
13-17
0
11-13
15-17
1,3
33-34
Taliev, 1955
Collocomcphorus
2
7-10
17-21
0
19-22
17-21
1,4
37-42
Taliev, 1955
Mclacotlus
1
7
19
0
13
16
1,3
—
Tahev. 1955
Paracottus
4
6-9
15-20
0
12-22
16-19
1.4
33-39
Taliev, 1955
Procottus
1
6-10
18-21
0
12-16
16-19
1.3
35-37
Taliev, 1955
Erenuniidae
Ereunias
1
9-11
12-15
0
11-13
14-15
0
36-39
Yabe. 1981; Watanabe. 1960. 1976
Marukawickthys
1
10
12-15
0
11-12
15
1,4
34-39
Yabe, 1981; Watanabe, 1960, 1976
Hexagrammidae
Hexagrammos
6
16-25
18-26
0-1"
19-26
17-21
1, 5
47-57
Kendall and Vinter, 1984; Washington
and Richardson, MS
Ophiodon
1
25-28
19-21
3**
21-25
16-18
1,5
56-59
Kendall and Vinter, 1984; Washington
and Richardson, MS
Pleurogrammus
1
21-24
24-30
0-1"
23-28
23-28
1,5
58-63
Kendall and Vinter, 1984; Washington
and Richardson, MS
Oxylehius
1
15-17
13-16
3-4**
14-17
14-17
1,5
36-40
Kendall and Vinter, 1984; Washington
and Richardson, MS
Zaniolepis
2
21-22
11-12
3"
18-20
14
1,5
40-43
Kendall and Vinter, 1984; Washington
Normanichthyidae
Normanichthys
10-11
11-12
14-15 17-19
I, 5
and Richardson, MS
36-37 Balbontin and Perez. 1980
444
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 112. Continued.
No. of
species
Dorsal fin
Ana! fin
Pecioral
fin
Pelvic
lin
Vertebrae
Genera
Spines
Rays
Spines
Rays
References
Psychrolutidae
Coltunculus
4
6-9
13-17
0
10-14
17-23
1.2-3
28-29
Howe-; Nelson, 1982
Dasycollus
1-2
8-11
13-16
0
12-16
22-26
1,3
34-35
Howe and Richardson. 1978;
Nelson, 1982
Ebinania
5
6-8
15-18
0
11-14
17-24
1.3
—
Nelson, 1982
Eurvmen
1
8
21-23
0
15-17
25-26
1.3
38
Howe and Richardson. 1978
*Maiacocoltus
4(?)
8-10
12-15
0
9-13
19-23
1.3
30-33
Howe and Richardson. 1978; Nelson,
1982
Neophrinichlhvs
2
7-12
14-18
0
11-14
23-26
1.3
31-34
Nelson. 1977
Psychrolutes
6
6-12
13-20
0
12-15
15-26
1.3
33-35
Nelson. 1982; Stein and Bond. 1978
* Taxonomic status is not agreed upon by current workers
•* No. of anal spines recognized vanes among workers.
' Howe. K. Compilation of menstic data from published and unpublished sources, available from Northwest and Alaska Fishenes Center. NOAA-NMFS, Seattle, WA.
- Howe. K.. unpublished data.
bladder is huge, occupying much of the body cavity with anterior
lobes reaching near the rear of the cranium. A more thorough
study of these gasbladders is needed. Dactylopterids. in relation
to triglids, have: 1) a hinged bony connection with the pre-
opercle; 2) much heavier and more elaborate ossification of the
cranium; 3) first three vertebrae elongate and modified; and, 4)
reduced opercular and gill openings. Given the extreme osteo-
logical modifications of these fishes, a current working hypoth-
esis (WNE) is that the Apistinae. triglids, peristediids and dac-
tylopterids share a common ancestry. However, more
information is needed before any formal changes are proposed.
Suborder Cottoidei
We include the following families in this group: Hexagram-
midae, Zaniolepidae. Anoplopomatidae. Cottidae (broad sense).
Agonidae. Cyclopteridae, and Liparididae. The limits of these
families and subfamilies are not well-defined and there is con-
siderable lack of understanding among workers in defining both
family limits and those of higher categories (see Washington
and Richardson, MS for review). We treat these diverse groups
together in order to facilitate discussion of past classifications,
not because we believe they necessarily form a monophyletic
assemblage.
Meristic features and approximate number of species for in-
cluded groups are provided in Table 1 1 2. Data have been com-
piled from many sources and may not be complete for some
groups and may omit extremes found in abnormal individuals.
Matsubara ( 1 955). in a thorough treatment of Japanese species,
recognized: 1) a superfamily "Hexagrammicae" (including An-
oplopomatidae and Hexagrammidae); 2) a superfamily "Cot-
ticae" (including Cottidae with subfamilies and Psychrolutidae);
and, 3) a superfamily "Agonicae" (including Agonidae and As-
pidophoridae). He placed the cyclopterids and liparidids in a
larger division, Cyclopterina.
Quast (1965), in a treatment which focused on relationships
of hexagrammid fishes, followed Regan (1913a) and Berg ( 1 940)
in recognizing a superfamily Hexagrammoidae and a superfam-
ily Cottoidae. He separated the Cottoidae from the Hexagram-
moidae on the basis of four characters: 1 ) lack of a basisphenoid;
2) dentigerous upper pharyngeals restricted to one or two pairs;
3) pleural ribs absent or developed on only a few posterior
abdominal vertebrae; and, 4) pectoral interradial foramina small
or absent. However, Quast proposed that the hexagrammids
and cottoids form a single evolutionary lineage within the Scor-
paeniformes and that the Anoplopomatidae are significantly
distinct from both the hexagrammid-cottid lineage and the scor-
paenid lineage to warrant separate superfamily status. He further
suggested that the zaniolepids are intermediate between the hex-
agrammids and cottids.
Other workers (Greenwood et al,, 1966; Nelson, 1976) have
placed the hexagrammids. anoplopomatids. and zaniolepids to-
gether in the suborder Hexagrammoidei. and the cottids (broad
sense of Washington and Richardson, MS), agonids, and cy-
clopterids in the suborder Cottoidei. Hallacher (1974) found a
cranioclavical (gasbladder) muscle present in the zaniolepids,
cottids (broad sense), agonids, and cyclopterids. In contrast,
Hexagrammos was found to have the scorpaenoid condition.
In the following discussion, we present information about
recent studies which have helped resolve relationships within
cottoid subgroups and outline groups where problems remain.
The systematic status of the Hexagrammidae is the subject
of disagreement at the specific through family levels. Quast (1965)
and Nelson (1976) include four genera in the Hexagrammidae—
Oxylehiits. Ophiodon, Hexagrammos (inc\u(i\n% Agramnms), and
Pleurogramfuiis. Quast considered Oxylehiits to be the most
primitive genus because of low numbers of meristic elements
and the "lack of specializations." Hexagrammos and Pleiiro-
grammus were considered to be closely related, relatively spe-
cialized genera because of the reduction in head spination. dorsal
and anal fin spines, etc.
Quast (1965) and Nelson (1976) included the two species of
the genus Zaniolepis in the family Zaniolepidae. Other workers
(Rutenberg, 1962) have included Zaniolepis in the family Hex-
agrammidae. while others (Hart, 1973) have combined Zani-
olepis and Oxylehius in the family Zaniolepidae.
The Anoplopomatidae contains two monotypic genera, An-
oplopoma and Erilepis (Quast, 1965; Nelson. 1976), however
some workers have placed Erilepis in its own family, the Eri-
lepidae.
Those families that traditionally have been placed in or near
the Cottidae are not clearly defined. Previous workers have
proposed between 1 and 1 7 families of "cottids." Greenwood
et al. (1966) and Nelson (1976) recognize 7 cottid families:
Cottidae, Icelidae. Cottocomephoridae. Comephoridae. Cot-
tunculidae, Psychrolutidae and Normanichthy idae. Other work-
ers have chosen to combine these 7 families in the single family
WASHINGTON ET AL.: SCORPAENIFORMES
445
Cottidae, until further study can define the phylogenetic rela-
tionships or monophyletic nature of these groups (Howe and
Richardson, 1978; Washington and Richardson, MS).
Yabe (1981) recognized the family Ereuniidae for the Jap-
anese "cottid" genera Ereunias and Marukawichthys. He used
derived characters such as free pectoral fin rays and associated
pectoral giidle modifications to define the family. Yabe con-
cluded that the genus hclus belonged in the Cottidae. Previous
workers (Matsubara, 1936; Berg. 1940; Nelson, 1976) have placed
Marukawichihys and Ereunias in the family Icelidae with mem-
bers of the genus Icelus.
Nelson (1982) has revised the "family' Psychrolutidae which
includes two subfamilies (Psychrolutinae and Cottunculinae).
Nelson could not define the family as monophyletic on the basis
of unique, derived characters and stated that the question of
whether to include the psychrolutids in the Cottidae was sub-
jective at this time. He rejected a close affinity between the
psychrolutids and liparidids as suggested by early workers.
The families Comephoridae and Cottocomephoridae are en-
demic to the Lake Baikal basin (U.S.S.R.). Berg (1940) recog-
nized each as separate families within the superfamily Cotto-
idae. Taliev (1955), after detailed study of the two groups,
suggested that they had originated from cottid ancestors and
cited as evidence their similarities to two cottid genera, Meso-
cottus and Trachydenmis. Both Taliev (1955) and Kozhov ( 1 963)
placed the cottocomephorids in the Cottidae while the vivipa-
rous comephorids were recognized as a separate family.
The family Agonidae has been reviewed only by Freeman
(1951) who suggested that the agonids were most closely related
to the cottids. The family is distinct in having fused, bony plates
covering the body.
Nonnanichthys crocken. the sole member of the Normanich-
thyidae, occurs off the coasts of Peru and Chile. Its relationships
are obscure. Norman (1938b) considered it to be a primitive
cottid, while others (Berg, 1940; Quast, 1965) have placed it in
its own family, in the superfamily Cottoidae. In addition to a
different body plan, the suborbital stay of Nonnanichthys is
quite distinct from other scorpaeniform fishes (Poss, 1975). Its
relationships to cottoids have yet to be established.
relatronships based on
Larval Characters
Larvae of only about 20% of the sc&rpaeniform genera are
known, and only recently have larvae been used in systematic
studies (see Richardson, 1981a; Washington, 1981; Kendall and
Vinter, 1984; Washington and Richardson, MS). The most ex-
tensive information dealing with systematic characters of scor-
paeniform larvae is presented in a recent study by Washington
and Richardson (MS). This work dealt with over 100 osteolog-
ical characters of larval and juvenile cottids and their allies.
About half of the 70 characters used in their analysis were re-
stricted to the larval period. In general, larval characters were
most useful in defining groups below the subordinal level.
Larvae of many scorpaenoid families are not yet known. Char-
acters such as head and preopercular spination and pectoral fin
length and pigmentation may be useful in future systematic
analyses; however, at present, larvae of too few taxa are known
to suggest relationships within the suborder Scorpaenoidei.
The results of Washington and Richardson's (MS) study, agree
with those of past studies which propose that a scorpaenid-like
stock was ancestral to the Scorpaeniformes and was derived
from a "generalized" perciform. Larvae ofthescorpaenid genera
Sebastes. Sebastolobus, and Scorpaena possess some characters
which are but slight modifications of those possessed by some
generalized percoids. In contrast, other scorpaeniform larvae
examined possessed considerable modifications of these char-
acters. These generalized scorpaenid characters include among
others: presence of predorsal bones; large, fused first anal pte-
rygiophore with three, stout anal spines; pleural ribs on abdom-
inal vertebrae; epipleurals attached to pleural ribs; hypurals 1 -I-
2 partially fused; hypurals 3-1-4 partially fused; presence of a
fifth hypural and parhypural; all hypural elements autogenous
and a specialized neural spine on preural centrum 2. Without
the suborbital stay, larvae of a scorpaenid such as Sebastes cou\d.
easily be mistaken for those of a generalized percoid. We con-
sider these character states to represent the plesiomorphic con-
dition in the Scorpaeniformes.
Washington and Richardson's study focused in detail on cot-
toid and hexagrammoid fishes where larvae of many taxa are
fairly well known. They found that the hexagrammoids exhibit
many character complexes which are derived relative to the
scorpaenids. These include: 1 ) reduced anal spines and first anal
pterygiophore; 2) the pleural and epipleural ribs inserted to-
gether on the vertebral parapophyses; and, 3) the pectoral radials
broadened and anvil-shaped, but with distinct foramina be-
tween them. None of these characters is unique to the larval
period.
Within the taxa traditionally assigned to the Hexagrammoidei
(Nelson, 1976), two monophyletic groups are recognized by
Washington and Richardson (MS). The first includes the hex-
agrammid genera Hexagrammos, Pleurogrammus. and Ophio-
don and the anoplopomatid genus Anoplopoma. This group is
defined by seven autopomorphies: 1) reduced head spination;
2) prolonged chondrification; 3) a unique (within Scorpaeni-
formes) sequence of ossification of the vertebral centra; 4) paired
first dorsal fin elements; 5) five preural centra involved in caudal
fin support; 6) anterior insertion of principal caudal rays; and,
7) a high number of vertebrae and ribs. The first four characters
are restricted to the larval period.
In contrast, larvae of the second group, Oxylcbius and Zan-
iolcpis, do not possess any of the synapomorphies of the first
group. They do share one derived character— an unfused neural
arch and spine of the first vertebral centrum. The arms of the
first neural "arch" and spine remain unfused for a brief time
during larval development, a unique condition among known
scorpaeniform larvae. Other larval characters support the sep-
aration of these groups, but we are cautious in the interpretation
of these characters. They include: 1) large versus small size at
hatching; 2) neustonic versus planktonic larvae; and 3) long,
slender versus deep body shape.
Washington and Richardson (MS) concluded that the first
group of hexagrammoids is very distinctive and differs from all
other scorpaeniforms so far examined, particularly in the mode
of ossification of the vertebral column and in the number of
preural centra involved in the caudal fin support. Because of
the uniqueness of these characters, Washington and Richardson
(MS) suggest that members of this hexagrammoid group prob-
ably comprise a separate lineage within the order, distinct from
Oxylebius and Zaniolepis and the other cottoids. The second
group, Oxylebius and Zaniolepis. is distinctive but appears to
be closer in many characters to the scorpaenids than to other
hexagrammoids.
446
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
<:^^
Fig. 241. Hypothesis of cottoid relationships modified from Washington and Richardson (MS).
Washington and Richardson's (MS) hypothesis of relation-
ships among the other cottoids studied is shown in Fig. 241.
Characters observed in the cottoid families, Cottidae (broad
sense of Washington and Richardson, MS), Agonidae, and Cy-
clopteridae, are derived relative to both the scorpaenids and
hexagrammids. The cottids, agonids, and cyclopterids share four
apomorphic characters, none of which is restricted to the larval
period. These include: 1) pleural ribs absent or restricted to the
posterior three abdominal vertebrae; 2) epipleurals independent
or sessile; 3) small first anal pterygiophore; and. 4) no anal
spines.
The cyclopterids (including lipandids) appear to be a distinct
family defined by a modified ventral sucking disc and are the
sister group of the cottids and agonids. (See Able, Markle and
Fahay, this volume, for discussion of cyclopterid relationships).
The cottids and agonids share three derived characters: 1 ) the
first anal pterygiophore is simple; 2) there are no supernumerary
anal elements; and 3) the haemal spine of preural centrum 2 is
enlarged. Again, none of these characters is unique to the larval
period.
Among the 28 genera of cottids examined, Washington and
Richardson (MS) recognized eight monophyletic groups which
are defined by one or more apomorphic characters. Rhampho-
cottiis, a monotypic genus, is characterized by four distinctive
autapomorphies, two of which are larval characters. Rhanipho-
cottus larvae possess a unique body shape with an extremely
long snout to anus length (>60% SL) and deep body shape (29-
40% SL). Rhamphocottus larvae also possess only one preoper-
cular spine. Other workers have also found Rhamphocoitits to
deviate from other cottids and have placed it in its own family
(Gill, 1888; Johnson, 19 18; Jordan, 1923;Bohn, 1934;Taranets,
1941).
Hemtlepidotus and Scorpaenichthys form another cottid group
and are defined by five autapomorphies, three of which are
unique to the larval period. First, members of both genera de-
velop heavy, pitted dermal bone on the cranium which forms
early in larval development. As the bone develops, ossification
proceeds unevenly with small pockets of bone apparently re-
sorbed forming pitted areas, while surrounding areas are thick-
ened. Second, larvae develop broad supraocular bony shelves
which project laterally over the orbit. Third, the dorsalmost
radial of the pectoral fin is reduced in size and becomes fused
or nearly fused to the scapula during larval development.
These three characters are not present in any other cottids
examined. Although both Scorpaenichthys and Hcimlepidotus
have been postulated as "primitive" cottids by workers studying
adults, they have not previously been considered closely related
to each other.
The remaining cottids and agonids share four additional de-
rived characters: 1) neural spine of PU 2 elongate; 2) neural
spine of first vertebra absent; 3) upper and lower hypural plates
fused to each other and fused to the urostyle; and. 4) the first
neural arch is unfused, rather it forms in a broad U-shape. The
last character is a larval feature found only in these taxa.
Five additional generic groups are defined by one to six au-
tapomorphies. Although these five groups contain the majority
of cottid genera, no synapomorphies were found which united
these groups and yet separated them from the agonids. The
WASHINGTON ET AL.: SCORPAENIFORMES
447
Myoxoccpha/us group includes 13 genera defined by the unique
larval character of a bony shelf on the anterior portion of the
preopercle. The Artedius group includes Clinocoiiiis. Oligocot-
liis. and Artedius Group A (see Washington, 1981). This group
is defined by six autapomorphic characters including three unique
larval features: 1) multiple preopercular spines; 2) enlargement
and expansion of the anterior neural arches; and, 3) first three
neural arches unfused. The Psychrolutes group includes Gilhert-
idia and Psychrolutes and is defined by six apomorphic char-
acters. Only one, the absence of head and preopercular spines,
is unique to the larval period. The Malacocottus group includes
Dasycottiis and Malacocottus and is defined by heavy, bony
arches on the cranium which form late in larval development.
Members of the last two groups were recently combined in the
family Psychrolutidae (Nelson, 1982) and correspond to his
subfamilies Psychrolutinae and Cottunculinae, respectively. The
Coitus group, including Cottus and Leptocottus. is defined by
four aulapomorphies, two of which are larval characteristics:
the first proximal dorsal pterygiophore is simple and slender in
contrast to all other cottid larvae and the parhypural is absent
in larvae of these genera. Further, larvae of these genera exhibit
a delay in ossification of the cranium and reduced head spi-
nation.
The last two cottoid groups are: the Hemitripterus group in-
cluding the "cottids" Hemitripterus, Nautichthys. and Blepsias.
and the agonids. These share three derived characters: 1 ) mod-
ified prickle-scales; 2) a knobby fronto-parietal ridge; and, 3)
broad plate-like epurals. The first two characters are unique
larval features of this group.
These characters provide evidence that the Hemitripterus
group, traditionally placed in the Cottidae, may be the sister
group of the Agonidae. Several agonid genera, such as Hypsa-
gomts and Agonomelas are very similar to members of the
Hemitripterus group both as larvae and adults. In addition,
larvae of these genera share several apparently derived char-
acters. However, the agonids, including Hypsagonus and Ago-
nomalus share several autapomorphies unique to the agonids
mcluding one or two plate-like epurals, and extreme modifi-
cations of the pectoral girdle.
The implications of these findings are that the agonids are
derived from the cottids and according to cladistic methodology
should be relegated to a sub-unit of the Cottidae. However,
Washington and Richardson (MS) do not propose any formal
changes in the cottids and agonids at this time. Larvae of only
about a third of the cottid genera have been studied. In addition,
the family or families of cottids have not been clearly defined
on the basis of derived characters, and until such time, we cannot
hope to fully understand the cottid-agonid interrelationships.
(B.B.W., K.M.H.) Gulf Coast Research Laboratory, East
Beach Drive, Ocean Springs, Mississippi 39564; (W.N.E.)
Department of Ichthyology, California Academy of
Sciences, Golden Gate Park, San Francisco, Califor-
nia 94118.
Tetraodontoidei: Development
J. M. Leis
THE tetraodontoid fishes (Gymnodontes) are a diverse sub-
order of one large and three small families and about 150
recent species (Winterbottom, 1974a; Tyler, 1980). The four
families (Table 1 13) are largely tropical, but many species are
temperate. Most species are marine and bottom-associated in
shallow waters, but the Molidae is entirely pelagic and both the
Diodontidae and Tetraodontidae have fully pelagic species. The
Tetraodonlidae also includes a number of fully freshwater species.
Many tetraodontoids have a pelagic, often oceanic, juvenile
stage.
Development
Development of tetraodontoid fishes is not particularly well-
known. Previous reviews of the early development of the group
are by Breder and Clark (1947), Tortonese (1956) and Martin
and Drewry ( 1978). Early development of triodontids is entirely
unknown, and, overall, information is available for only 36
species. The information available for particular species of these
36 is often scanty. However, for the Molidae, information is
available for all three species. Complete (i.e., egg to juvenile)
information is available for about 10 species (Table 1 14). In the
following sections, I assume that the few taxa for which infor-
mation is available are representative (these taxa and the de-
velopmental stages concerned are listed in Table 1 14). The fol-
lowing sections should be read in conjunction with Table 1 14;
citations listed in Table 114 are not repeated in the text. In
parentheses after the family heading I give the number of species
for which some information is available.
On the basis of early life history characters, the tetraodontoid
fishes are a more coherent group than the balistoid fishes.
Table 113. Merlstic Characters of Tetraodontoid Fishes
Principally after Tyler (1980). N is the approximate number of
recent species largely after Nelson (1976). Pelvic fins are lacking in this
suborder.
Family
N
C
Vene-
brae
Triodontidae
Tetraodontidae
Diodontidae
Molidae
I
130
15
3
0-n, 1 1
7-34
10-18
15-20
10
7-27
10-18
14-18
15-16
12-20
18-25
7-13
12
II
9-10
12-26*
20
16-30
18-23
16-18
* Not a true caudal fin. but a clavus or pseudocaudal.
448
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 14, Tetraodontoid Taxa for whk h Information is Available on Egg and Larval Stages. YS — yolk-sac stage; pre — preflexion
stage; flex — flexion stage; post — postflexion stage; U — unstated; den- demersal; pel — pelagic; PS— examined for the present study. Numbers in
parentheses after each genus refer to the number of species represented. A blank means no information available on that stage.
Eggs
Larvae
developmental stage
Type
Oil
droplets
Tetraodontidae
Canthigasler ( > 1 )
dem
0.68-0.72
cluster
x
Carmoletraodon ( 1 )
U
0.78
U
Chelonodon ( 1 )
dem
U
U
Chonerhinos (5)
U
1.1-2.3
U
Fugu (7)
dem
0.85-1.32
cluster
x
Lagocephalus ( 1 )
dem
0.61-0.70
cluster
X
Sphoeroidcs ( 1 )
dem
0.85-0.91
cluster
X
Tetraodon ( 1 )
dem
1.0
U
Torquigener (2)
dem?
0.94
U
X
Unidentified (7)
X
Diodontidae
Allomvcterus
pel
2.0-2.2
20-25
Chilomycterus (3)
pel
U-1.8
U
Diodon (3)
Tragiilichlhys ( 1 )
Molidae-
Masliirus ( 1 )
Mola ( 1 )
Ranzania ( 1 )
pel
1.62-2.1
cluster
pel 1.8 cluster
pel 1.42-1.68 cluster
X Fujita, 1962; Stroud et al., MS; PS
Breder and Rosen, 1966
Breder and Rosen, 1966
Roberts, 1982a
X Uchida et al., 1958; Mito, 1966; Masuda et
al., 1975; Fujita. 1962
X Uchida et al., 1958; Fujita, 1962. 1966
X Welsh and Breder. 1921
Breder and Rosen, 1 966
X Munro, 1945
X Miller et al., 1979; Leis and Rcnnis. 1983;
PS
Robertson, 1975a
Evermann and Kendall, 1898; Breder, 1927;
Nichols and Breder, 1927; Fowler, 1945;
Breder and Clark, 1947; Heck and Wein-
stein, 1978; Moyer, pers. comm.; Fujita.
1962; PS
Sanzo, 1930d'; Mito. 1966; Leis, 1978; Sak-
amoto and Suzuki, 1978; Fujita, 1962; PS
PS
X Schmidt, 1921; Martin and Drewry, 1978
X Martin and Drewry, 1978
X Schmidt, 1921; Leis. 1977
' Misideniified as Crayracion sp. (Tetraodontidae).
■ No caudal fin forms: pre, fiex and post in this case refer to clavus formation, not notochord lleMon,
£gg.j.— Tetraodontoid fishes are oviparous. Pelagic and demer-
sal eggs are known; the chorion is smooth; clusters of oil droplets
are present; eggs range in size from large (2. 1 mm) to small (0.6
mm) and are spherical; incubation times are long and range
from 3 to 20 days; development at hatching varies; the peri-
vitelline space is narrow; the yolk is unsegmented; and embryos
may be heavily pigmented. Parental care of eggs is present only
in some tetraodontids.
Larvae. — M\ tetraodontoid larvae are pelagic. Development in
most tetraodontids is direct; in molids and diodontids special-
ized ontogenetic stages may exist. There are few larval special-
izations except in the Molidae, and development is usually com-
pleted at a small size. There is often an apparently unspecialized
pelagic juvenile stage, which may be very large at settlement.
Larvae are enclosed in a more or less inflated vesicular sac.
Larvae are deep and wide in head and trunk, and the tail is
comparatively small and compressed. The head is large and
rounded and the gut is coiled and massive. The eye is partic-
ularly large. The specialized adult scales form directly (i.e., do
not pass through an unspecialized spinule stage). In molids spe-
cialized larval spines are formed. The pectoral fin is the first to
form, and the caudal fin is last. Except for the tail of molids,
structures are not formed and subsequently lost — they never
form. The specialized dentition develops during the larval stage
directly, without any intervening generalized teeth. However,
diodontids and tetraodontids may have small, raised points
along the cutting edges of their beak-like teeth. Meristic char-
acters are summarized in Table 1 1 3 [see Tyler ( 1 980) for further
information]. The number of vertebrae is low ( 16-30), as is the
number of caudal fin rays (0-12). Pelvic fins arc lacking and
except for some triodontids. the fins lack spines. Larvae are
heavily pigmented. The few larval specializations which do oc-
cur are the vesicular dermal sac of all species and the huge
dermal spines of molids.
Only two groups have specialized ontogenetic stages between
larvae and juveniles. In the Diodontidae, some Atlantic species
of Chilomycterus (sensu lato) have a postflexion stage {"Lyo-
sphaera") that lacks dermal spines, but has fleshy protuberences
in the locations the spines will occupy and other enlarged pro-
tuberences unassociated with spines (Evermann and Kendall,
1898; Breder, 1927; Heck and Weinstein, 1978). In the Molidae,
Mola and Mastwus have a deep-bodied, compressed stage
("Molacanlhus") that has reduced larval spines, and a distinctly
non-adult shape (Martin and Drewry, 1978).
Family Accounts
Triodontidae. — 'Ho\.hm% is known of triodontid eggs or larvae.
Tetraodontidae. — Tetraodontid eggs are demersal, small to me-
dium-sized, have multiple oil droplets (Table 1 14) and hatch
in 3-20 days. The very large ovarian eggs of Chonerhinos (Table
1 14), a highly specialized freshwater genus (Tyler, 1980; Rob-
erts, 1982a), are here regarded as a specialization for freshwater
LEIS: TETRAODONTOIDEI
449
Fig. 242. Tetraodontoid yolk-sac larvae. All specimens are enclosed
in a more or less well-developed vesicular dermal sac. The vesicles are
omitted in the drawmgs. From top to bottom: Lagocephalus tunaris
(Tetraodontidae) 1.7 mm (1.9 mm TL) (after Fujita, 1966); Fugii par-
dalis (Tetraodontidae) 2.6 mm (2.84 mm TL) (after Uchida et al., 1 958);
Diodon (hystnx^) (Diodontidae) 2.6 mm (after Leis, 1978); and Ran-
-aiiia laevis (Molidae) 1.8 mm (after Lets, 1977).
conditions [freshwater species commonly have larger eggs than
their marine confamiliais (Roberts, pers. comm.)]. The chorion
is adhesive. Parental care of eggs is known, but not universal.
Development of larvae at hatching varies with species: jaws
totally unformed to partially formed; the eye ranges from un-
pigmented to completely pigmented; the pectoral bud may be
present or absent; a moderately developed vesicular dermal sac
encloses head and trunk; much yolk remains; and pigment ranges
from moderate to heavy (Fig. 242). If the often huge yolk sac
is ignored, larvae are initially cylindrical, but become progres-
sively deeper and wider-bodied with growth (Fig. 243). Larvae
Fig. 243. Tetraodontoid larvae. From top to bottom: Unidentified
tctraodontid larva (possibly Canlhigaslcr), 3.6 mm, from the Great
Barner Reef. Note small spines in skin; Tragnlichlhys jacitliferus (Dio-
dontidae), 4.2 mm. from the Great Barrier Reef (small circles in the
dermal sac represent incipient spines and arc ossified); and Ranzania
lac'vis (Molidae), 3.9 mm (after Leis, 1977).
remain deeper than broad until they acquire the ability to inflate.
Until mid-preflexion stage the body remains relatively fusiform
with a well-developed tail (relative to other tetraodontoids and
ostraciids). The moderately-developed vesicular sac often dis-
appears during the pretlexion stage, but may be retained in some
species until after flexion. This sac does not correspond to the
inflatable belly found in this family. The gill opening closes to
a pore shortly after the yolk is absorbed, but the membranes
are thin and transparent and thus easily missed. Sequence of fin
450
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
formation is: P.-D = A-C'. The long notochord tip persists for
a time following flexion. The vesicles of the dermal sac are said
to be the source of the small dermal ossifications (Welsh and
Breder, 1921). The dermal ossifications (=scales) develop di-
rectly into small, often embedded, spines. The dermal spines
seem to first appear on the belly, usually in the preflexion stage.
Depending on species, the spines may appear on the rest of the
body shortly thereafter, slowly and gradually, or not at all. Pig-
ment is mitially heavy over the gut, brain and yolk sac, and
usually spreads to cover much of the head and trunk before
flexion. Welsh and Breder (1921) and Munro (1945) report the
presence of a single opercular spine in preflexion larvae of
Sphoeroides maculalus and Torquigener pleurogramina, re-
spectively. None of the larvae examined for the present study
has such a spine, but these two species have not been examined.
Diodontidae. — Diodontid eggs are pelagic, large, have multiple
oil droplets (Table 114) and hatch in 3 to 5 days. Larvae are
moderately to well developed at hatching, but development var-
ies between species and possibly between populations of the
same species: jaws range from totally unformed to formed and
apparently functional; eyes are partially to fully pigmented; the
gill opening is reduced to a pore; moderate to heavy pigment
(including yellow, red and orange) is present; much yolk re-
mains; and a well-developed, inflated, vesicular dermal sac en-
closes head and trunk (Fig. 242). Larvae are deep-bodied and
broader than deep (Fig. 243). At hatching or very shortly there-
after, diodontid larvae are extremely rotund with head and trunk
a single ball-like unit. The tail is small and becomes relatively
smaller still with age. It becomes nearly vestigial during flexion,
but thereafter starts to increase in size. Body shape changes little
during development. The fins form P.-D = A-C. The mouth
is large compared with other tetraodontiform larvae. Shortly
before flexion, lens-like thickenings form in the dermal sac, and
(depending on species) small swellings or elongate papillae form
over these. The large spines (=scales) subsequently form inside
these structures without an intermediate stage. In most species,
spines are present around the time flexion is completed, but in
Chilomyclerus antennatus and C. schoepfi (but not C. affinis or
C. orbicularis) there is a specialized pelagic stage which lacks
' Sequence of ossification of first element in each fin, except that the
symbol for caudal fin (C) refers to completion of notochord flexion. Fin
preceding dash forms prior to fin following dash.
spines and may have some of the elongate papillae enormously
enlarged (the genus Lyosphaera was described from such a stage).
The spines in the "Lyosphaera" stage form after settlement.
Nostrils of diodontids form in a conventional manner. Only
following development of a short tentacle with two openings do
the split nasal flaps of Dicotylichthys or the open reticulated
nasal cups of Ctiilomyclerus ajfinis form during the late juvenile
stage. Pigment is moderate to heavy and in preflexion larvae
much heavier dorsally than ventrally. Following flexion, there
is a tendency for the belly to become more heavily pigmented
than the dorsum.
A/olidae. — Mo\id eggs are pelagic, large, have multiple oil drop-
lets (Table 114). and hatch in 7 to 8 days. Larvae are devel-
opmentally very advanced at hatching with: jaws formed; eyes
pigmented; gill opening reduced to a pore; a well-developed
vesicular dermal sac enclosing head and trunk; the cleithrum
and several pectoral fin rays ossified; a dorsal fin anlage; heavy
pigment; and an unknown amount of yolk (Fig. 242). The body
is deep (Fig. 243) and wide but not as wide as deep. At hatching
molid larvae are extremely rotund with head and trunk a single
ball-like unit. The compressed tail becomes progressively small-
er. With growth and body spine development the body even-
tually becomes more compressed and a ventral keel forms. The
fins form Pi-D = A-Clavus. The P, forms very early and be-
comes large. The tail of young larvae is normal, but soon begins
to atrophy, and a true caudal fin never forms. Notochord flexion
does not take place, so the clavus is not homologous with the
caudal fin. Shortly after hatching, the huge spines which char-
acterize molid larvae begin to form. These reach a maximum
size at about the time the clavus is formed. As the massive
spines decrease in size, small spines form elsewhere, particularly
on the ventral keel. Also, small ossifications within the skin
begin to form, and these eventually make up the carapace-like
skin covering. Mo/a and Mastwus pass through a fairly long
ontogenetic stage between larvae and juveniles which is char-
acterized by retention of reduced massive spines, a deep, com-
pressed body with a ventral keel and a shape quite unlike the
adult (the genus Molacanthus was described from such a stage).
Ranzania. in contrast, loses its spines relatively quickly and
directly assumes the adult shape. Larvae are heavily pigmented
over the gut and on the dorsal surfaces.
Department of Ichthyology, The Australian Museum, P.O.
Box A285, Sydney, 2000, Australia.
Balistoidei: Development
A. Aboussouan and J. M. Leis
THE tetraodontiform suborder Balistoidei (Sclerodermi) is a
small group of six families with about 175 recent species
of great morphological diversity (Tyler, 1968, 1980; Winter-
bottom, 1974a;Matsuura, 1979).Thesuborderisgenerallyagreed
to consist of the six families (Table 115) considered here (Tyler,
1980). However, Winterbottom (1974a) has suggested that the
triacanthodids and triacanthids could be removed to a suborder
distinct from all other tetraodontiform fishes. The group is large-
ly tropical and marine, but some species range well into the
temperate zones, particularly in Australia. Most species are bot-
ABOUSSOUAN AND LEIS: BALISTOIDEI
451
Table 1 15. Range of Mi
ERISTIC
Characters
OF Balistoids.
Mostly after Tyler,
1980: see this and Matsuura, 1979 for further information.
Tnacanthodidae
Triacanthidae
Balistidae
Monacanlhidae
Aracanidae
Ostraciidae
Number of species
19
7
35
90
12
15
(after Nelson, 1976)
Dorsal spines
6
4 or 6
3
1 or 2
0
0
Second dorsal rays
12-18
19-26
23-35
22-50
9-12
9-11
Anal rays
11-17
13-22
19-31
20-62
9-11
8-11
Pectoral rays
12-15
12-16
12-17
8-16
10-13
9-13
Caudal rays
6 + 6
6 + 6
6 + 6
6 + 6
5 + 5-6
5 + 5
Pelvic spines
1
1
0
0
0
0
Pelvic rays
0-2
0-1
0
0
0
0
Ventral scales
absent
absent
present
absent or
present
absent
absent
Vertebrae
8
+ 12 = 20
8+12 = 20
7 + 10-12 =
17-19
6-8 + 1 1-23 =
19-31
9-
-10 + 8-9 =
18
9 + 9-10 =
18-19
Caudal fin bones
Epural
1
1
1
1
0
0
Uroneural
2
1
1 ?
0
0
0
Hypural
3 to 5
3
3
2 or 3
2
2
Smallest hypural (5th)
present
present
present
absent or
present
absent
absent
Parhypural
1
1
1
1
0
0
Vertebrae before the
8
4-5
5
4-7
5-6
6-8
first second dorsal
pterygiophore
Vertebrae behind the last
5-6
7
4-5
4-6
4-6
4-6
anal pterygiophore
tom-associated in shallow to moderate depths, but many tria-
canthodids live in deep (>500 m) water. Most species have a
pelagic, often oceanic, juvenile stage, and a few are pelagic
throughout their lives.
Development
Development of balistoid fishes is not well known. Previous
reviews of the early development of the group are by Breder
and Clark (1947), Tortonese (1956). and Martin and Drewry
(1978). The early development of aracanids is entirely unknown,
and. overall information is available for only 30 species. Often
the information available for a species is scanty. Complete (i.e.,
egg to juvenile) information is available for only four or five
species (Table 1 16). This narrow data base makes generaliza-
tions about development somewhat suspect. However, we as-
sume that the few taxa for which information is available are
representative.
Few generalizations can be made about development of balis-
toid fishes, but this is not surprising in view of the diversity of
the adults. A reference to development in juveniles, which usu-
ally differ little from adults, is given at the end of each section.
We make no attempt to review the literature on juvenile de-
velopment.
Eggs (Table //6A — Balistoid fishes are oviparous. Pelagic and
demersal eggs are known: the chorion is usually smooth, but
may have limited sculpturing; oil droplets are usually present;
eggs range in size from small (0.5 mm) to large (2.0 mm) and
are approximately spherical; incubation times range from one
to four days; development at hatching vanes widely; the peri-
vitelline space is narrow; the yolk is unsegmented; and consid-
erable pigment may develop on the embryo. Parental care of
eggs ranges from non-existent (pelagic eggs) to considerable (Ba-
listidae).
Larvae— AW balistoid larvae are pelagic. Development is gen-
erally direct (i.e., no specialized ontogenetic stages between lar-
vae and juveniles), with few larval specializations, and is com-
pleted at a small size (Figs. 244-25 1 ). There is often an apparently
unspecialized pelagic juvenile stage which may grow to a sig-
nificant fraction of the adult size. Larvae tend to be deep-bodied,
and many are also wide-bodied. The head is large and the gut
coiled and massive. The mouth is small. The head is usually
rounded, at least in preflexion larvae. The head and body of
young ostraciid larvae are enclosed in an inflated dermal sac
which has numerous vesicles (or tubercles) embedded in its
outer surface. Except in ostraciids, the specialized adult scales
pass through an unspecialized spinule stage. The caudal fin is
usually the last fin to form. The reduction in structures, notably
fins, which characterizes the balistoid fishes is not a case of
development followed by loss— these structures never develop.
The specialized dentition develops during the larval stage di-
rectly, without any intervening generalized or larval teeth. Me-
ristic characters are summarized in Table 1 15 (see Tyler. 1980
for further information). The number of vertebrae and caudal
fin rays is low. pelvic fins are reduced or lacking, anal fins lack
spines, and dorsal spines, if present, are few (Table 1 1 5). Larvae
are generally moderately to heavily pigmented.
The few larval morphological specializations which do occur
are either developments of the often very specialized scales (or
their precursors) of the adults or delicate skin flaps, filaments
and tendrils. These are discussed under the appropriate family
section. There are no specialized ontogenetic stages between
larvae and juveniles.
A shorthand notation will be used to designate the sequence
of fin formation. By formation, we mean ossification of the first
element, with the exception of the caudal fin where completion
of flexion is meant. However, except for some monacanthids
452
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 16. Balistoid Taxa for which Information is Available on Egg and Larval Stages. YS— yolk-sac stage; pre — preflexion stage;
flex— flexion stage; post — postflexion stage; U — unstated; dem— demersal; pel— pelagic. Numbers in parentheses after each genus refer to the
number of species represented. A blank means no information available.
Eggs
Lar\'ae
developmental stage
Type
Size
(mm)
Oil
droplets
YS
Pre
Rex
Post
Triacanthodidae
Atrophacanthus ( 1 )
Macrorhamphosodes ? ( 1 )
X
X
X
X
Triacanthidae
Triacanthus ( 1 )
Unidentified (1)
pel
0.78
0
X
X
X
X
X
X
Balistidae
Balisles (2)
dem
U
1
X
X
X
X
Balistapus ( 1 )
Canthidermis (1-2?)
dem
dem
0.55
U
0?=
u
X
X
X
X
Odonus ( 1 )
Pseudobalisles (3)
dem
dem
U
0.55-0.60
u
0?-
Sufflamen (3)
dem
0.51-0.56
1 & u
X
X
Xanrhichlhys ( 1 )
Unidentified (> 1)
Monacanthidae
Alulera (2)
A manses ( 1 )
Anacanlhus (1)
Brachaluteres ( 1 )
Camheriiies ( 1 )
Monacanlhus ( I )
Navodon ( 1 )
Parika ( 1 )
Pseudalutaris ( 1 )
Rudanus ( 1 )
Stephanolepis (3)
Unidentified (>10)
Ostraciidae
Acanlhoslracion ( 1 )
Lac lor ia (2)
Ostracion (2)
Tetrosomus (1)
Unidentified (6)
Rhmesoinus (1)
dem
pel' 0.65-0.74' 1'
dem 0.52 2
dem 0.61-0.70 cluster X
pel 1.4-1.6 1 X
pel 1.6-1.9 cluster X
pel 1.6-1.9 cluster X
pel
1.4-2.0
cluster
X
X X
X X
X
X
X X
X
X X
X X
X
X
X
X
X
X
X
X
X
X
X
Fraser-Brunner, 1950; Tyler, 1968; present study
Present study
Ohsima and Nakamura, 1941; Gopinath, 1946
Present study
Sanzo, 1939b'; Gamaud, 1 960; Aboussouan, 1966;
Lythgoe and Lythgoe, 1975; Matsuura and
Katsuragawa. 1981
Lobel and Johannes, 1980
Nellis, 1980; Watson and Walker, pers. comm.;
present study
Fricke, 1980
Masuda et al., 1975; Fricke, 1980; Lobel and
Johannes, 1980; Matsuura, 1982
Ballard, 1970; Sweatman, pers. comm.; Thresher,
pers. comm.; present study
Present study
Leis and Rennis, 1983; present study
Clark, 1950; Suzuki et al., 1980; present study
Present study
Present study
Leis and Rennis (1983, figure 72); present study
Present study
Present study
Uchida et al., 1958; Kobayashi and Abe, 1962;
Mho, 1966
Regan, 1916; Robertson, 1975a; Crossland, 1981
Lets and Rennis, 1983; present study
Fujita, 1955; Uchida et al., 1958; Kobayashi and
Abe, 1962; Mito, 1966; Masuda et al., 1975;
Suzuki etal., 1980
Ryder, 1887; Hildebrand and Cable, 1930; Fujita,
1955; Uchida et al.. 1958; Mito, 1966; Aboussouan,
1966
Leis and Rennis, 1983; Watson and Walker, pers.
comm.; present study
Brederand Clark, 1947; Palko and Richards, 1969;
present study
Watson and Leis, 1974; Moyer, 1979; Leis and
Meyer, MS; present study
Watson and Leis, 1974; Leis and Rennis, 1983; Leis
and Moyer, MS; present study
Present study
Delsman, 1930d-'; Sanzo, 1930d'; Mito, 1962c, 1966;
present study
Present study
Notes:
' Two specimens (L86 and 2.48 mm) from a supposed senes of Dactvluplerus volnans appear to be Baiisla capmcus. and one has the preopercular cluster of spinules. A 4 mm specimen
identified as B capnscus is also illustrated.
' Lobel and Johannes (1980) descnbe the eggs as "without visible inclusions." but their photograph of a newly hatched B. imdulalus seems to show an oil drop in the yolk sac.
' Eggs identified as Wovodon [sic] convexiroslris i= Parika scaber) were described by Robertson (1975a) and Crossland ( 1981 ). however there is reason to question this identification- Robertson
(in lill- Nov. 1982) notes the identification and classification as pelagic of this egg was "based on a small sample of npe |unfcnili/cd] eggs from a female leatherjacket and a conforming type in
the Olago Harbour plankton at that time," and that no eggs were reared. We feel the eggs described by Robertson and Crossland are not monacanthids.
* Misidentified as Tclraodon sp.
■ Misidentified as Tclraodon honkcnn
ABOUSSOUAN AND LEIS: BALISTOIDEI
453
Fig. 244. Scanning electron micrograph of the sculptured chorion
of an unidentified Hawaiian ostraciid egg. The micropyle is the hole in
the center. The width of the field of bumps is cci. 0.5 mm.
Fig. 245. Triacanthid and ostraciid yolk-sac larvae from top to bot-
tom: Triacanthus hiaculeatus. 1.3 mm ( 1 .4 mm TL) after Ohsima and
Nakamura. 1941; and Acanthostracion quadricornis. 2.6 mm reared
larva from Rorida. Specimen is fully enclosed in a vesicular sac which
is most inflated over head and trunk. The vesicles are omitted in the
drawing. Specimen is unpigmented, but is probably bleached.
where the posterior rays of dorsal and anal fins are slow to form,
ossification of all elements of the fin could serve as an equally
good definition. The fins will be indicated by standard notation
(D— dorsal, Dsp— dorsal spine, etc.). The order of the letters
corresponds to the order of formation. An equal sign between
two letters indicates the fins form simultaneously, a dash in-
dicates the fins do not form simultaneously.
Triacanthodidae
The eggs of triacanthodids are unknown, although there is a
dubious report of pelagic eggs (Nikol'skii, 1961). The body of
preflcxion and flexion larvae (Fig. 247) is moderately to very
deep, moderately wide in head and trunk, and compressed in
tail. The body becomes more compressed and elongate with
growth, but may remain very deep until well after flexion. The
gill opening is closed to a pore in the smallest available speci-
mens (late preflexion). There is no dermal sac. The fins form
D = A = P,-C-P, = Dsp. The Dsp anlage and P, buds do not
form until after flexion. Although no early postflexion larvae
are available, late flexion larvae have a notochord with a long
posterior portion that probably indicates that the notochord has
an extended tip for awhile following flexion. Dermal spinules
first form in preflexion larvae, and appear first on side of head
(cheek, operculum, over otic vesicle) and laterally on two small
regions of the gut (ventral to P, base and just anterior to anus).
The spinules are unspecialized, and fully cover the body of
postflexion larvae. The available larvae of Atrophacanthus are
unpigmented, but their poor condition implies they could be
faded. The Macrorhamphosodcs (?) larva is moderately and uni-
formly pigmented with small melanophores.
The specimen identified as Triacaruhodes sp. by Weber (1913)
appears to be a trichiurid (Scombroidei), not a triacanthodid.
Tyler (1968) describes juvenile development of several tria-
canthodid species.
Triacanthidae
Triacanthid eggs lack oil droplets and chorion sculpture, are
pelagic, small, and hatch in about 22 hours (Table 1 16). De-
velopment at hatching is not advanced (Fig. 245): no jaws or
pectoral fins are present, the eye is unpigmented and much yolk
remains. The body is cylindrical at hatching and becomes much
deeper with growth (Fig. 247) and, especially following flexion,
very compressed. The gill opening closes to a pore prior to
flexion. There is no dermal sac. The fins form D = A = P|-P; =
Dsp-C. The notochord has an extended tip following flexion.
The D and P, spines become relatively elongate. Dermal spi-
nules first form in preflexion larvae and appear first on the sides
of the head (cheek, operculum, over otic vesicle), and laterally
on the posterior portion of the gut. The spinules are unspecial-
ized (except for some terete ones on the fin spines), and fully
cover the body shortly after flexion. Pigment is heavy on brain
and gut. and a single ventral tail melanophore is present. Fol-
lowing yolk exhaustion, pigment spreads over most of the body
in a blotchy pattern.
454
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 246. Balistid and monacanthid yolk-sac larvae from top to
bottom: Sufflainen chrysopterus. 1.7 mm reared larva from the Great
Barrier Reef (24 hours after hatching); Stephanolepis cirrhtfer. 1.9 mm
(2. 1 mm TL) after Fujita, 1955; and Anacanthns harbalus. 2.4 mm larva
from a Great Barrier Reef plankton sample (age unknown). Note anlage
of dorsal fin spine in occipital region. Mouth is not fully formed. Frag-
mented oil droplets are present in the yolk sac. but are not illustrated.
The descriptions of lai^ae identified as Triacanthus breviros-
tris by Kuthalingam ( 1 959b) do not resemble triacanthid larvae
in morphology, sequence of development, or size. One can only
conclude the larvae are misidentified and the drawings inac-
curate. The eggs identified by Kuthalingam (1959b) as Tria-
canthus brevirostris are probably those of an atheriniform fish.
Tyler (1968) describes juvenile development of several tria-
canthid species.
Bahstidae
Balistid eggs are demersal, small, lack chorion ornamentation
(but are adhesive), have a single oil droplet (Table 1 16), and
hatch in one to two days. Eggs are laid in clusters in shallow
nests on sand or rubble bottoms and are guarded by the adult.
Development of larvae at hatching is not advanced: no jaws are
present, the eye is unpigmented, minimal body pigment is pres-
ent and much yolk remains (Fig. 246). Larvae have a cylindrical,
slightly compressed body at hatching. The body quickly be-
comes deeper and then moderately rotund in the trunk (Figs.
248 and 249). The tail remains compressed. About the time fins
start to form, the larva starts to become compressed and this
increases thereafter. In newly hatched larvae, a slightly inflated
area is present surrounding the trunk (Fig. 246), but it contains
no vesicles, and soon disappears. The gill opening closes to a
pore just prior to flexion. The fins form Dsp-D = A = P,-C.
Fig. 247. Late preflexion larvae of three balistoid families. Small
ticks on upper and middle specimens indicate position of dermal spi-
nules. From top to bottom: Atrophacanlhus japonicus (Triacanthodi-
dae), composite drawing of three damaged larvae (2.6-2.7 mm) from a
Dana Station in the Philippines; unidentified triacanthid. 3.5 mm, from
the Great Barrier Reef note small dorsal spine and pelvic fin bud; and
Acanlhostracion QuaiJricornis (OsUaciidae). 3.3 mm. reared larva from
Florida. The dermal plates arc not yet formed, but ridges on the body
are evident.
The first dorsal spine becomes large and heavily armed with
barbs before flexion. This ornamentation varies between species
and is useful in identification. The notochord has an extended
tip for a short while following flexion. Dermal ossifications first
ABOUSSOUAN AND LEIS: BALISTOIDEI
455
Table 117.
Characters That Differ Between the Two Larval
monacanthid morphs.
Fig. 248. Late to mid preflexion larvae of two balistoid families.
Small ticks indicate position of dermal spinules. From top to bottom:
Canlhidenms sufflamen (Balistidae), 3.5 mm, from Puerto Rico, note
small pelvic bud and prcopercularclustcr of spinules; unidentified Morph
A monacanthid, 3.6 mm, from the Great Bamer Reef, note pigmented
filament at terminus of pel vie bone and preopercular cluster of spinules;
Pseudatulans nasicorms (Morph B monacanthid), 4.3 mm, from the
Great Barrier Reef, note pigmented fleshy tendrils laterally on tail and
preopercular cluster of spinules; unidentified Morph C monacanthid,
3.0 mm from the Great Bamer Reef Dermal spinules in this species
are longer than in the other illustrated species. Dorsal spine is just
beginning to form (not yet ossified).
Character
Morph
AB (Figs. 246 and 248)
C (Fig. 248)
Body shape
Deep to elongate; be-
Deep; becoming
coming angular
deeper with
with growth. Com-
growth, but re-
pressed.
maining rounded.
Somewhat rotund
early, becoming
compressed.
Cluster of spinules
Small to large
Absent
on preoperculum
Sequence of fin for-
Dsp-D = A-P,-
D = A = C-P,-
mation
or = C
Dsp
Dorsal fin spine
Early-forming, armed
Late-forming, lightly
or unarmed, lightly
pigmented
to moderately pig-
mented
Tail pigment in
Present. Ventral or
None
preflexion larvae
dorsal series or
blotches.
Identified taxa in-
Alutera. Amanses.
Brachaluteres. Ru-
cluded
Anacanthus.
danus (the morph
Cantherines. Mon-
C larvae illustrated
acanthus, Navo-
by Leis and Rennis
don. Parika, Pseu-
(1983, Fig. 72) are
dalularis.
Brachaluteres)
Stephanolepis
appear in the form of a small cluster of relatively long spinules
on the preoperculum (larval specialization). This cluster appears
within a few days of hatching and persists until just prior to
flexion. Shortly before the cluster disappears, dermal spinules
appear in three areas: laterally on the cheek ventral to the cluster;
over the otic vesicle; and laterally on the gut from below the
pectoral base to near the anus. These unspecialized spinules
rapidly spread to cover the body by mid-flexion. They do not
transform into the specialized scales of the adults until well into
the pelagic juvenile stage. A pigmented filament (larval spe-
cialization) often develops at the terminus of the pelvic bones
(see discussion of such structures under Monacanthidae). Pig-
ment is heavy on the brain and gut, and preflexion larvae have
a series of melanophores on the ventral midline of the tail.
Blotches or bands may form on the tail. The spiny dorsal fin is
heavily pigmented and this pigment spreads laterally over the
trunk.
Berry and Baldwin (1966) describe juvenile development of
several balistid species.
Monacanthidae
Monacanthid eggs are demersal (we tentatively conclude that
pelagic eggs were wrongly attributed to Parika scaher—see Table
1 16), adhesive, small, have several oil droplets, and hatch in
about 2 days. Eggs are attached to vegetation, and there is no
record of parental care. Development at hatching is not ad-
vanced: jaws are absent or only partially formed, the eye is
unpigmented, and much yolk remains (Fig. 246). Newly-hatched
larv ae are cylindrical and somewhat compressed. Morphological
and developmental diversity among monacanthid larvae is high.
456
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
■BSCor-
Fig. 249. Balistid preflexion larva Xanlichthys ringens 3.87 mm, ASFIOl, western Atlantic.
particularly in comparison with other tetraodontiform families.
Leis and Rennis (1983) considered that three distinct morphs
were present among larval monacanthids. However, our studies
of additional taxa indicate that Morphs A and B of Leis and
Rennis (1983) are merely extremes (Fig. 248), and that no clear
division can be made between A and B. For example, while Leis
and Rennis (1983) utilized seven characters to separate the two
morphs, larvae of Alutera sp. (Fig. 250) have three 'A' char-
acters, three 'B' characters, and are intermediate for the seventh.
Morph C of Leis and Rennis (1983) (Fig. 248) is distinct from
the combined Morph AB (Table 1 17). Rapid changes in body
proportions may take place in many species. Morph AB larvae
are compressed and become more so with growth, while Morph
C larvae are moderately broad in gut and head, but become
compressed with growth. The gill opening closes to a pore late
in the preflexion stage, and the position of the pore relative to
the eye varies with species. The dorsal spines form at a very
early stage in Morph AB larvae, but are the last fin elements to
form in Morph C larvae. In some species, the first spine becomes
heavily armed by the mid-preflexion stage. The pelvis may form
either early and be prominent by the mid-preflexion stage or
very late and may never become externally visible, depending
on species. The sequence of fin development is morph-depen-
dent (Table 1 17). There is no dermal sac. The notochord tip is
long and persists for a time following flexion. If present (Morph
AB), the small cluster of spinules on the preoperculum (a larval
specialization) forms very early and is lost before flexion. De-
pending on species, dermal spinules may first appear laterally
on the gut and head, or on the forehead and along the ventral
midline near the cleithral symphysis. Dermal spinules cover the
body prior to flexion or shortly thereafter. Pigment is heavy on
the brain and over the gut, but on the tail, it varies with species.
Several species temporarily develop small, pigmented flaps
or filaments (a larval specialization) on different portions of the
body. Alutera sp. (Fig. 250) develops an elongate flap which
originates on the operculum near the preopercular spinule clus-
ter; Pseudaliiiaris nasicornis (Fig. 248) develops several, elon-
gate tendrils laterally on the tail; and many species develop a
filament at the terminus of the pelvic bones (Fig. 248). The
latter possibly represents a pelvic fin bud that atrophies.
The description of Slephanolepis hispidus by Hildebrand and
Cable ( 1 930) seems to be based on more than one monacanthid
species (Martin and Drewry, 1978). Berry and Vogelc (1961)
describe the juvenile development of several monacanthid
species.
Hildebrand and Cable (1930) state that the pelvis oi' Sleph-
anolepis hispidus possibly forms through coalescence of two
separate fin buds. In the material available to us (Table 1 16),
ABOUSSOUAN AND LEIS: BALISTOIDEI
457
^tS'i"
/::;::.
y.^.f
Mor-
Fig. 250. Monacanthid preflexion larvae, (upper) A manses pulhis 3.99 mm SL, Y-92; (lower) ,-l/M;era sp. 2.76 mm SL, ASF-94 western Atlantic
(preopercular cluster of spinules not seen in these specimens).
the pelvis forms from a single unpaired aniage located just pos-
terior to the cleithral symphysis. The terminal encasing scales
form first as unspecialized spinules, and at the same time a pair
of pelvic elements begins to ossify. The pelvis then fuses begin-
ning from its base (e.g., in a 2.35 mm specimen, the pelvis is
roughly 'Y'-shaped and fused along about 75% of its length).
The "two ventral fins" observed by Hildebrand and Cable (1930)
are probably dermal flaps similar to those of Alutera sp.
Development of the preopercular cluster of spinules, aside
from Leis and Rennis (1983), has been described in published
works only for Batistes capnscus (Sanzo. 1939b; Matsuura and
Katsuragawa, 1981) and Pahka scaber (Crossland, 1981), al-
Fig. 251. Ostraciid larvae from top to bottom: Rhmesoinus triqueter 2.85 mm SL, ASF-37, western Atlantic; Laclophrys quadncornis 2.53
mm SL, and 6.0 mm SL ASF94, western Atlantic (exhalent gill openings not shown).
ABOUSSOUAN AND LEIS: BALISTOIDEI
459
though in the latter it is illustrated as a serrate preopercular
border. However, this structure is present in all balistids and
Morph AB monacanthids examined for the present study, and
because it is an inconspicuous structure it is most likely that it
is present in previously described taxa and has been overlooked
(see Fig. 250).
Aracanidae
Nothing is known of aracanid eggs or larvae.
Ostraciidae
Ostraciid eggs are pelagic, large, slightly ovoid, have one or
more oil droplets (Table 1 16) and hatch in two to four days.
There is some chorion ornamentation surrounding the micro-
pyle. In Indo-Pacific species (Ostraciinae) this consists of a par-
tially raised field of small bumps surrounding a small pore-like
depression containing the micropyle (Fig. 244). In the single
Atlantic species examined (Lactophrysinae), only the pore-like
depression is present. Development of larvae at hatching is rel-
atively advanced, but there is some interspecific variation in
how advanced: jaws are totally unformed to formed and ap-
parently functional, the eye is unpigmented to partially pig-
mented, dorsal and anal fin anlagen may be present. Moderate
pigment is present, much yolk remains, the gill opening is re-
stricted to a pore, and an inflated vesicular dermal sac encloses
head and trunk (Fig. 244). The dermal sac disappears before
flexion. The larvae are deep-bodied and the tail is compressed
(Fig. 247). Depending on species, the body may be moderately
(Rhinesomus) to very wide (Ostracion) (Fig. 251): the lacto-
phrysine species examined were more narrow-bodied than the
ostraciine species. Larvae tend to become wider with growth,
but never become as wide as deep. At hatching, ostraciine larvae
are rotund with head and trunk a single ball-like unit, and they
have a small tail. Lactophrysine larvae attain this condition
within a few days of hatching. The tail progressively becomes
relatively smaller with age until after flexion, and the ball-like
shape of the body is retained. The notochord tip is small. The
lips have an unusual flared structure. The fins form P.-D = A-
C. The dermal ossifications do not pass through a spinule stage,
but form directly starting as thickenings in the dermal sac which
ossify and grow out from their centers. These eventually coalese
into the mosaic-like armoured carapace characteristic of adults.
The individual carapace units that eventually produce spines
and other ornamentation tend to be larger and with more relief
than other carapace units. The ossifications become visible well
before flexion, and larvae are fully armoured by the end of
flexion. Pigment is moderate to heavy and generally uniform
on head and trunk, with the tail often unpigmented.
Le Danois (1961) describes the juvenile development of sev-
eral ostraciid species.
Chorion ornamentation of ostraciid eggs previously has been
reported only for Hawaiian taxa (Watson and Leis, 1974; Leis,
1977, 1978), however it is present in all ostraciid eggs examined
in the present study (Fig. 244), albeit reduced to a pore in Acan-
thostracion quadncornis (Table 1 1 6). The ornamentation is sub-
tle and confined to a small portion of the chorion, and we feel
it is probably present in all taxa, but has been overlooked in
previous descriptions.
(A. A.) Station Marine D'Endoume et Centre D'Oceanog-
RAPHiE, Rue Batterie des Lions, 13007, Marseille,
France; (J.M.L.) Department of Ichthyology, The
Australian Museum, P.O. Box A285, Sydney, N.S.W.,
2000, Australia.
Tetraodontiformes: Relationships
J. M. Leis
IN this contribution I construct a phylogeny of tetraodontiform
fishes based on early life history (ELH) characters and con-
trast this with phylogenies based on adult characteristics. The
ELH characters of tetraodontiform fishes are summarized in the
preceding two contributions (Aboussouan and Leis, and Leis,
this volume). Although in many cases there is little information
available, I have assumed that which is available is represen-
tative, and that new information will not change the conclusions
herein. This is unlikely, and for this reason, the present treat-
ment must be viewed with caution.
Inter-ordinal Relationships
The tetraodontiform fishes are usually presumed to have been
derived from perciform ancestors, with the Acanthuroidei being
the popular choice for closest relative (Tyler, 1980; Winterbot-
tom, 1974a; Lauder and Liem, 1983). However, D. E. Rosen in
an unpublished study (pers. comm.) presents evidence sup-
porting a relationship between zeiform and tetraodontiform fishes
(see also Winterbottom, 1974a).
There is little in the early life history of tetraodontiform fishes
to indicate they are the sister group of the acanthuroid fishes.
The few ELH characters acanthuroids and tetraodontiforms share
(small mouth, gas bladder present, relatively few myomeres,
large head, oviparity, spherical eggs with unsegmented yolk) are
very widespread in the perciform fishes, and the larvae are not
even generally similar (see Leis and Rennis, 1983). Certain char-
acter states (e.g., scale development) are shared by the acanthu-
roid fishes and some groups of tetraodontiform fishes. This sit-
uation could be interpreted as indicating a relationship between
acanthuroids and tetraodontiforms, whereupon the character
state involved would be viewed as primitive for the Tetra-
odontiformes as a whole. Therefore, the presence of an alternate
character state in some tetraodontiform families would be viewed
as a derived condition. This type of interpretation, while prob-
ably realistic, is avoided here as it is fraught with opportunities
for circular reasoning.
Too little is known of ELH characters in zeiform fishes (Tighe
and Keene, this volume) to enable a proper evaluation of the
460
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Zeiformes as a potential sister group to the tetraodontiform
fishes. However, at least 7.eus and Capros have a long notochord
tip; Capros has early-forming spinule-like scales; and Capros
and Zeus (but not Antigoma) larvae are generally similar to
some balistoid larvae in body shape and pigmentation. Thus,
based on scanty ELH information, there are some suggestions
of support for Rosen's proposal of a zeiform-telraodontiform
relationship.
Present knowledge of ELH characters does not help much in
determining the inter-ordinal relationships of the tetraodonti-
form fishes. This is partially because there are no unique, derived
ELH characters shared by all tetraodontiform fishes (see below).
In addition the ELH characters which are shared between te-
traodontiforms and either acanthuroids or zeiforms (above) are
shared with other groups as well, thus lessening the value of
these characters in determining relationships: e.g., similar body
shape, pigment, reduced number of vertebrae, early-forming
spinule-like scales, and elongate notochord tip are found in var-
ious combinations in priacanthids, pomacanthids, callionymids
andlophiiform fishes (Leis and Rennis, 1983, and relevant chap-
ters in this volume).
Therefore, one must rely on ideas of inter-ordinal relation-
ships based on adults. For the purposes of this analysis, the
Acanthuroidei and the Zeiformes are considered as alternative
sister groups for the Tetraodontiformes.
Order Tetraodontiformes
There are relatively few ELH characters which apply to the
Tetraodontiformes as a whole, and fewer still which could be
considered derived. The only characters which might be con-
sidered derived are the late formation of the caudal fin and the
various early-forming scale specializations, and both are found
in a few other percoid and non-percoid groups. The dermal sac
and some other derived characters are probably derived within
the order and are of no use in characterizing the order as a
whole. Other tetraodontiform characters which are wide-spread
among other fishes are: small mouth, gas bladder present, rel-
atively few myomeres and fin rays, large head, no bones of the
head with spines, oviparity, basically spherical eggs with un-
segmented yolk, and transformation to an unspecialized pelagic
juvenile at a small size. Therefore, I could find no uniquely
derived ELH characters shared by all members of the order.
Some features of the adults can be considered paedomorphic:
large head, lack of certain structures that simply never form
(Fraser-Brunner, 1950), delayed ossification of some bones.
Intra-ordinal relationships
As noted above, the Acanthuroidei (consisting of the families
Acanthuridae, Zanclidae and Siganidae) and the Zeiformes (in-
cluding Caproidae after Rosen, pers. comm.) will be considered
as alternative sister groups to the Tetraodontiformes. Therefore,
characteristics shared with the early life history stages of the
Acanthuroidei, and particularly the Acanthuridae (or alterna-
tively with the Zeiformes) will be considered primitive. Char-
acteristics of acanthuroid larvae are summarized in Leis and
Rennis (1983) and Leis and Richards (this volume). Character-
istics of zeiform larvae are summarized in Russell (1976) and
Tighe and Keene (this volume).
Two tetraodontiform families cannot be included for lack of
information (Aracanidae and Triodonlidae) and these are not
considered further. I don't know how seriously these omissions
might bias the results. It is assumed the egg characteristics of
the triacanthodids (which are unknown) are the same as those
of the triacanthids.
Perhaps surprisingly, the acanthuroid and zeiform character
states differ for only three of the characters used in the following
analysis. For these three, the difference lies in my inability to
assign polarity to the character if the zeiforms were chosen as
the outgroup. Thus, it makes no difference to the shape of the
resulting phylogeny (but does weaken two of the branch points)
if the Zeiformes rather than the Acanthuroidei is chosen as
outgroup.
A discussion of the characters used follows (Table 118): (1)
Egg type— Acanthurids (and zeiforms) have pelagic eggs, al-
though siganids have demersal eggs. The demersal eggs of te-
traodontiform fishes and siganids have no adaptations for being
demersal other than stickiness or a mucous mass, and seem
relatively unspecialized for being demersal. A pelagic egg is
considered primitive. (2) Egg size— Acanthuroid eggs are small
(< 1 mm), so eggs larger than 1.4 mm are considered derived.
However, zeiform eggs are medium to large (0.95-2.0 mm), so
if zeiform fishes are accepted as the outgroup, polarity of this
character cannot be determined. (3) Oil droplets in eggs— Acan-
thuroid eggs (and zeiform eggs) have one or more oil droplets
in the yolk. Lack of oil droplets in eggs is considered derived.
(4) Egg shape— An egg that is not spherical is considered derived
because acanthuroid eggs (and zeiform eggs) are spherical. (5)
Chorion sculpture — Sculpturing on the chorion is considered
derived because acanthuroid eggs (and zeiform eggs) are un-
sculptured. (6) Incubation period— Acanthuroid eggs hatch in
about two days or less, and an incubation time longer than this
is considered derived. Because incubation period is tempera-
ture-dependent, it is possible that some of the differences noted
here are artifacts of the different temperatures at which the eggs
were reared. However, insofar as it has been possible to com-
pare different taxa reared at similar temperatures, the differences
in incubation period noted here seem valid. Incubation times
of zeiform eggs are poorly known, but may be up to 13 days for
Zeus. Therefore, if zeiform fishes are accepted as the outgroup,
polarity of this character cannot be determined. (7) Parental
care of eggs— There is no parental care of eggs by fishes with
pelagic eggs including zeiforms and acanthurids. Siganids lay
demersal eggs but no parental care has been reported. Therefore,
lack of parental care is considered primitive. (8) Body shape—
Acanthuroid (and zeiform) larvae tend to be cylindrical to some-
what compressed at hatching and to be compressed by the time
flexion is complete, although they may pass through an early
preflexion stage which is more rotund. This developmental pat-
tern is considered primitive. Some tetraodontiform larvae are
extremely rotund throughout development, but this is largely
due to a greatly inflated dermal sac (see character 10). (9) Head
and gut development — All balistoid fishes but ostraciids hatch
with a cylindrical to compressed body. All of these but mona-
canthid Morph AB become deeper-bodied and wider in head
and gut by the middle of the preflexion stage and then become
compressed by flexion. Morph AB monacanthids never become
broad in head and gut. Due to the widespread occurrence of the
wide body development mode in the suborder, it is considered
primitive. (10) Vesicular dermal sac— Some tetraodontiform
larvae have the head and trunk enclosed in a vesicular dermal
sac, a condition not found in acanthuroids or zeiforms (a very
weakly-developed dermal sac without vesicles similar to the one
of yolk-sac balistids is found in acanthurids). This sac and its
subdermal space seem to be the source of many of the dermal
LEIS: TETRAODONTIFORMES
461
Table 1 18. Early Life History Characteristics of the Tetraodontiform Fishes. (P) indicates primitive, and (D) derived. (?) indicates
assumed, (s) indicates that character is secondarily m state given. ( — ) indicates not applicable for family. See text for discussion of characteristics.
(*) indicates character for which polarity cannot be established if the Zeiformes is considered the sister group of the Tetraodontiformes.
Taxon
Mona-
Mona-
Tnacantho-
Tnacanthi-
Bahsti-
canthidae
canthidae
Telra-
Diodonli-
Characler
didae
dae
dac
A B
C
Ostraciidae
odontidae
dae
Molidae
1.
Egg type
P?
P
D
D
D
P
D
P
P
*2.
Egg size
P?
P
P
P
P
D
P
D
D
3.
Oil droplets
D?
D
P
P
P
P
P
P
P
4.
Egg shape
P?
P
P
P
P
D
P
P
P
5.
Chorion sculpture
P?
P
P
P
P?
D
P
P
P
*6.
Incubation period
P?
P
P
P
P?
D
D
D
D
7.
Parental care of eggs
P?
P
D
P
P?
P
P-D
P
P
8.
Body shape
P
P
P
P
P
D
D
D
D
9.
Head and gut development
P
P
P
D
P
—
—
—
—
10a.
Vesicular dermal sac
P
P
P
P
P
D
D
D
D
10b.
Dermal sac inflation
—
—
—
—
—
D
P
D
D
11a.
Opercular pore A
P?
P
P
P
P
D
D
D
D
lib.
Opercular pore B
—
—
—
—
—
D
P
D
D
12.
Scale development
P
P
P
P
P
D
D
D
D
13.
Very large spines
P
P
P
P
P
P
P
P
D
14.
Preopercular cluster
P
P
D
D
Ps
P
P
P
P
15.
Long notochord tip
P
P
P
P
P
P
D
D
D
16.
Dorsal spine development
D
D
P
P
Ds
—
—
—
—
17a.
Dorsal spines A
P
P
D
D
D
D
D
D
D
17b.
Dorsal spines B
P
P
P
D
D
—
—
—
—
17c.
Dorsal spines C
P
P
P
P
P
D
D
D
D
18.
Dsp, P,sp formation
D
P
P
P
P
—
—
—
—
19.
Pelvic fin
P
P
D
D
D
D
D
D
D
20.
Pelvis
P
P
P
P
P
D
D
D
D
*21a.
Caudal fin rays (£ 1 1)
P
P
P
P
P
D
D
D
D
21b.
Caudal fin rays (s 10)
P
P
P
P
P
D
P
D
D
21c.
Caudal fin rays (none)
P
P
P
P
P
P
P
P
D
22.
Pectoral development
P
P
P
P
P
D
D
D
D
23.
Body width
P
P
P
P
P
P
P
D
P
specializations of the Tetraodontiformes. A dermal sac is con-
sidered derived (10a). A strongly inflated dermal sac (with a
large subdermal space) linking head and trunk in a ball-like unit
is considered a further derivation from the presence of a sac
(lOb). (II) Restriction of gill opening to a pore — In perciforms
with restricted gill openings (no zeiform fishes have restricted
gill openings), the opercles are fully open in early larvae and
gradually close to a pore. The assumed primitive condition in
Tetraodontiformes is for closure to a pore to occur after some
rays of the median fins have ossified (11a) because this is closest
to the perciform condition. Having the opening closed to a pore
at hatching is considered derived from closure at the end of the
yolk-sac stage (lib). (12) Scale development— The specialized
scales of adult tetraodontiform fishes form in two ways: directly
or by first passing through a relatively unspecialized spinule
stage. The intermediate spinule stage is considered primitive
because it is present in acanthurids and at least Capros in the
zeiforms. In acanthurids small spine-like scales change into tri-
angular scales arrayed in vertical rows, and these spine-like
scales form first on the lower gut and laterally on the head, in
the same place they first form in most tetraodontiform fishes
which have them. In Capros small spinules form prior to flexion
and eventually cover the whole body before differentiating into
scales. (13) Very large dermal spines (larval specialization)—
The ver>' large, fixed spines with conical or pyramidal bases of
molids are unique and are considered derived. ( 1 4) Preopercular
spinule cluster (larval specialization)— This cluster of spinules
is unique to balistids and most monacanthids and is considered
a derived character. (15) Long notochord tip— The notochord
may extend well past the caudal fin aniage, and if so, following
flexion it will protrude dorsal and parallel to the caudal rays (to
about '/: their length) for a time. This condition was initially
considered derived because it is absent in acanthuroid fishes.
However, in the Tetraodontiformes the long notochord tip is
absent only in taxa in which the tail becomes greatly reduced
(i.e., Diodontidae, Ostraciidae, Molidae). Therefore, it seems
better to regard the long notochord tip as a primitive characler
within the order, but a character derived after the supposed split
from the acanthuroid fishes. The absence of this structure within
the order is thus derived. Zeiform larvae (Zeus. Capros) have
an elongate notochord tip very similar to that of balistoid fishes,
so this is considered the primitive condition. (16) Dorsal fin
spine development sequence— Dorsal fin spine development
prior to dorsal fin soft ray development is considered primitive
because this is the condition in acanthuroids and zeiforms. ( 1 7)
Dorsal fin spines— Acanthuroid fishes have 4 to 14 dorsal fin
spines, and zeiform fishes 5 to 10. Therefore, in the tetraodon-
tiform fishes, the most primitive character state is the greatest
number of spines (i.e., 4-6 of triacanthodids and triacanthids).
The intermediate derived condition is a reduction in this num-
ber to three spines ( 1 7a). From the intermediate condition are
derived one or two spines ( I 7b) and no spines ( I 7c). ( 1 8) Initial
formation of fin spines— The presence of dorsal fin spine aniage
and pelvic fin buds prior to flexion in fishes that have late-
462
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
<
Q
UJ
<
Q
J-
U
o
2
<
<
1-
O
Q
O
Q
_1
O
o
Q
Z
13.21c
2.IOb,llb.l5.2lb
TETRAODONTOIDEA
6. 8.10a. I la. 12. 17c, 20. 21a. 22
TRIACANTHOIDEI
TETRAODONTOIDEI
TETRAODONTIFORMES
Fig. 252. Phylogeny of tetraodontiform fishes based on early life history characters (excluding Aracanidae and Triodontidae for which no
information is available). Numbers refer to characters (see text) and are located on the branch which possesses the derived state. Characters 2, 6
and 21a would be deleted if the Zeiformes is accepted as the sister group of the Tetraodontiformes. Note that character 1 occurs m two places
indicating conflict with the accepted classification.
forming dorsal spines (character 16) is considered primitive
because it is closer to the presumed ancestral (i.e., acanthuroid
and zeiform) condition of early-forming dorsal and pelvic spines
than is late formation oftheanlage. (19) Pelvic fin — Acanthuroid
(and zeiform) fishes have a pelvic fin formula of at least I, 3.
Presence of a pelvic fin is primitive, and its absence is derived.
(20) Pelvis— The pelvis is present in acanthuroid (and zeiform)
fishes, and its absence is considered derived. (21) Caudal fin
rays- Acanthuroids have 16-17 principal caudal fin rays. The
maximum number (i.e., 1 2) in tetraodontiform fishes is consid-
ered primitive. The intermediate derived condition is < 1 1 rays
(2 1 a). The next most advanced condition is < 1 0 rays (21b), and
the most advanced condition is the complete absence of the
caudal fin (21c). Zeiform fishes have 1 1-15 principal caudal fin
rays, so if zeiforms are accepted as the outgroup, polarity of 2 la
cannot be determined, while 21b and 21c would not change.
(22) Pectoral fin development— The pectoral fin in acanthuroid
(and probably zeiform) fishes develops after or simultaneously
with the dorsal and anal fin soft rays, and this is considered
primitive. (23) Body width — The condition of body width >
body depth found in the Diodontidae is unique in the Tetra-
odontiformes and is considered derived.
Phylogenetic Analysis
Relationships within the Tetraodontiformes based on ELH
characters are presented in Fig. 252. In the following section, I
will contrast the present phylogeny with three phylogenies based
on adult characters (Fig. 253): myology (Winterbottom, 1 974a),
external and internal characters (Tyler, 1980) and osteology
(Rosen, pers. comm.). Lauder and Liem's (1983) review of in-
terrelationships of tetraodontiform fishes depends heavily on
Winterbottom's (1974a) work and, for my purposes here, is
identical to his phylogeny. Therefore, Lauder and Liem's (1983)
phylogeny will not be considered separately. The ELH-based
phylogeny exactly matches none of the three adult-based schemes,
but is closest to Rosen's (pers. comm.), differing only in place-
ment of the Tetraodontidae. Two cautions should be kept in
mind: 1) Rosen's (pers. comm.) study is primarily concerned
with inter-ordinal relationships, and the portion dealing with
intra-ordinal relationships of the Tetraodontiformes is based on
relatively few characters; and 2) the present phylogeny has lim-
itations flowing from exclusion of two families and many
subfamilial taxa due to lack of information.
There is most agreement between the four phylogenies in the
question of the relationship of the triacanthodids (Figs. 252,
253). The present phylogeny and those of Winterbottom (1974a)
and Rosen (pers. comm.) agree in the erection of the suborder
Triacanthoidei as the sister group to all other tetraodontiform
fishes. Tyler (1980) includes the triacanthodids in the balistoid
line, but this is a result of philosophy of classification more than
anything else (Tyler, pers. comm.).
The four phylogenies are evenly divided on the question of
LEIS: TETRAODONTIFORMES
463
TYLER
WINTERBOTTOM
I TRIACANTHODIDAE
2 TRIACANTHIDAE
3 BALISTIDAE
4 MONACANTHIDAE
5 OSTRACIIDAE
6 TETRAODONTIDAE
7 DIODONTIDAE
8 MOLIDAE
Fig. 253. The two published adult-based phylogenies of tetraodon-
tiform fishes which were tested by the ELH-based phylogeny. These
phylogenies were modified by omitting the two families which could
not be included in the ELH-based phylogeny. After Tyler (1980) and
Winterbottom (1974a). Numbers refer to the families listed at bottom.
Rosen's unpublished phylogeny is not shown.
ostraciid relationships (Figs. 252, 253), indicating that further
study is required. Winterbottom ( 1 974a) and Tyler ( 1 980) place
the ostraciids with the balistoids, a view reinforced by a recent
reassessment of their data (Winterbottom and Tyler, 1983). The
present phylogeny and Rosen's (pers. comm.), however, place
the ostraciids with the tetraodontoids. A relationship between
ostraciids and tetraodontoids was suggested by Sakamoto and
Suzuki (1978) based on general similarity of larvae.
The three adult-based phylogenies regard the tetraodontids
as the sister group of the diodontids (Fig. 253). This differs
significantly from the ELH-based phylogeny (Fig. 252) which
regards the tetraodontids as the sister group of all other tetra-
odontoids (including ostraciids). The trichotomy between these
"other tetraodontoids" in Fig. 252 cannot be resolved at present.
Further study is indicated.
The balistoids (Monacanthidae-Balistidae) branch off in a
convincing manner, but not without problems. The phylogeny
as depicted in Fig. 252 requires that demersal eggs (1) be in-
dependently derived in balistoids and tetraodontids. Although
this is quite possible, it brings into question the validity of using
demersal eggs as a derived character to define the Balistoidea.
Morph C monacanthids lack the preopercular spine cluster (14)
which characterises all other balistoids. I conclude that this is
a secondary loss and that the delayed development of the dorsal
fin spine in Morph C is independently derived (thus not indi-
cating a relationship with triacanthoids).
All phylogenies agree on the close relationship of monacan-
thids and balistids. Indeed, in the present study (Fig. 252), they
were separated by only two ELH characters, (17b) loss of a fin
spine, and (7) parental care of eggs, about which there is little
information and which is variable in tetraodontids. Although
the present phylogeny is nominally consistent with Matsuura's
(1979) phylogeny, Winterbottom (1974a) considered monacan-
thids and balistids to be subfamilies, and the ELH-based phy-
logeny presented here has done little to clarify this conflict.
There is some indication from ELH characters of divergences
within families, but the amount that can be said is severely
limited by the small number of taxa for which ELH characters
are known. The diodontids seem very conservative but some
species of Chilomycterus have a specialized ontogenetic stage
between larvae and juveniles (" Lyosphaera"): this supports re-
moval of these species to a separate genus (study in progress).
Within the ostraciids, the two subfamilies are separated by de-
gree of chorion ornamentation, and to a lesser degree by de-
velopment at hatching. The specialized "Molacanthus" stage
separates Afola and Masturus from Ranzania in the Molidae.
Balistids seem very conservative in development. Tetraodontids
vary greatly in development at hatching, parental care of eggs,
and perhaps in a number of other characters. Too few taxa are
known within the triacanthodids and triacanthids for any state-
ments to be made here. Monacanthids have the most variation
in ELH characters within the order, some of which has already
been referred to (Aboussouan and Leis, this volume). There
seems to be a great deal of potential in the use of ELH characters
for phylogenetic studies in the Monacanthidae, but first, devel-
opmental series for more species and genera must be established.
I have attempted to use ELH characters independently as a
test of phylogenies based on adult characters. Where the two
types of phylogenies support each other, confidence in the phy-
logeny is increased. Where differences appear, further study, or
re-interpretation of existing data is called for to resolve the
differences.
In conclusion, the present classification should be viewed with
caution because there are relatively few taxa for which eady life
history information is available. Monophyly of the tetraodon-
tiform fishes could not be established using ELH characters.
The present ELH-based phylogeny and those of Winterbottom
(1974) and Rosen (pers. comm.) agree in the creation of a sep-
arate suborder for triacanthoid fishes; Tyler (1980) disagrees
with this placement. Tyler (1980) and Winterbottom (1974a)
agree in placing the Ostraciidae in the Balistoidea, in contrast
to inclusion of the Ostraciidae within the Tetraodontoidea as
proposed here and by Rosen (pers. comm.). My placement of
the Tetraodontidae is in conflict with previous phylogenies based
on adult characters. In other areas, the ELH-based phylogeny
is in agreement with the three adult-based phylogenies. The
different placements of the Tetraodontidae and in particular the
Ostraciidae in the present classification warrant further inves-
tigation of tetraodontiform interrelationships.
Department of Ichthyology, The Australian Museum, P.O.
Box A285, Sydney, N.S.W., 2000, Australia.
Percoidei: Development and Relationships
G. D. Johnson
AS the largest and most diverse of the perciform suborders,
the Percoidei exemplifies the inadequacies that charac-
terize perciform classification. Regan (1913b) defined the Per-
coidei "by the absence of the special peculiarities which char-
acterize the other suborders of the Percomorphi [=Perciformes],"
and seventy years of research in systematic ichthyology have
failed to produce a more meaningful definition. In the absence
of even a single shared specialization uniting the percoids, the
monophyly of this great assemblage of fishes is doubtful. In spite
of our inability to adequately define the Percoidei, or because
of it, half of the approximately 145 families of perciform fishes
are usually referred to this suborder. Greenwood et al. (1966)
listed 71 percoid families in their "highly tentative" familial
classification of the Perciformes, and Nelson ( 1976) stated that
the Percoidei contains 72 families, 595 genera and about 3,935
species.
Percoids are best represented in the nearshore marine envi-
ronment and form a significant component of the reef associated
fish fauna of tropical and subtropical seas. A few groups are
primarily epipelagic or mesopelagic. Association with brackish
water occurs in many nearshore marine families, some of which
have one or more exclusively freshwater members, but only
four families are primarily restricted to freshwaters, the north
temperate Percidae and Centrarchidae, the south temperate Per-
cichthyidae (with one brackish water species) and the tropical
Nandidae.
In a practical sense, the suborder Percoidei serves the Per-
ciformes in much the same capacity as the Serranidae once
served the Percoidei itself as a convenient repository for those
"generalized" perciform families that cannot obviously be placed
elsewhere. I have treated the percoids in a similar sense here,
one of practicality and convenience. 1 do not intend to imply
or formulate hypotheses about the monophyly of the Percoidei
or to consider their intrarelationships as a whole. My major
objectives are to provide some preliminary documentation of
the variability of a number of character complexes among adults
and larvae of those fishes we now call percoids, to suggest what
1 believe to be promising avenues of future research and to offer
some specific examples illustrating the utility of larval mor-
phology in elucidating percoid phylogeny.
Classification
As here defined (Table 1 1 9) the Percoidei includes 80 families
and 1 2 incertae sedis genera, making it by far the largest and
most diverse suborder of teleostean fishes. The overall limits of
the suborder are only slightly modified from Greenwood et al.
(1966). The Pomacentridae, Embiotocidae and Cichlidae are
excluded because of their recent placement in the Labroidei by
Kaufman and Liem (1982). The suborder Acanthuroidea is
treated separately in this volume, but a recent hypothesis (Mok
and Shen, 1983), with which 1 concur, based on additional evi-
dence, suggests a close relationship between acanthuroids and
the Scatophagidae. The affinities of the questionably monophy-
letic Nandidae remain unresolved (Lauder and Liem, 1983),
and although the nandids are provisionally included in my list
of percoid families, they were not considered in the larval and
adult tables. The genus Elassoma. formerly a member of the
family Centrarchidae, is excluded from the Percoidei, for rea-
sons discussed below. The monophyly of the suborder Trachi-
noidei, as defined by Greenwood et al. (1966) is suspect, and
the affinities of families such as the Mugiloididae, Percophidae,
Chiasmodontidae and others may lie with the percoids. How-
ever, these families are treated elsewhere in this volume, and
of the "trachinoids," only the Opistognathidae are here included
as percoids.
Although the overall limits of the Percoidei are similarly per-
ceived in my classification and that of Greenwood et al. (1966),
substantive discrepancies result from differences in concepts of
family limits. For example, my Serranidae (Johnson, 1983) in-
cludes the Pseudogrammidae and Grammistidae of Greenwood
et al. (1966). Leptohrama is treated as a monotypic family sep-
arate from the Pempherididae (Tominaga, 1965), epigonids are
treated as a separate family, etc. The high percentage of mono-
typic families that has historically characterized percoid clas-
sification is a disturbing but unavoidable problem that can only
be remedied with a better understanding of percoid intrarela-
tionships. In my classification (Table 1 19), 26 of the 80 families
are monotypic and 12 genera, which lack family names, are
retained incertae sedis. Families and incertae sedis genera are
arranged alphabetically for easy reference and to avoid any in-
ference of affinity based on sequence. The classification of
Springer (1982) was followed for most families treated by him
and otherwise that of Nelson ( 1 976). Below, 1 discuss differences
between my classification and that of Spnnger ( 1 982) or that of
Nelson (1976), and present some new information about fa-
milial relationships. Early life history information contributed
substantially to some of these modifications.
Acropomatidae and Symphysanodon— The "oceanic per-
cichthyids" of Gosline (1966) do not share the defining char-
acteristics of the Percichthyidae (see below), and are treated here
as a separate family, including the following genera— v^cropowa,
Apogonops. Doederleinia (=Rhomhoscrranus), Malakichthys,
Neoscombrops. Synagrops and V'erilus. I know of no synapo-
morphy that unites the acropomatids, and further work will be
necessary to test their monophyly. Larvae of four genera are
known. Those of Acropoma (Fig. 254C), Doederleinia (Fig. 254D)
and Malakichthys are quite similar, but those of Synagrops {Fig.
254B) differ in pigmentation, body form, and the presence of
more extensive head spination. Although the larvae of Sym-
physanodon (Fig. 254A) are unique in their possession of horn-
like frontal spines, they are otherwise remarkably similar to
those of Synagrops (Fig. 2548), suggesting that these two genera
may be closely related.
Callanthiidae and Grammatidae.— Springer (1982) noted that
"there is little evidence to unite" the five genera he included in
the family Grammatidae. I concur with this and treat two of
these genera, Callanthias and Grammatonotus as a distinct fam-
ily, the Callanthiidae (currently under revision in collaboration
464
JOHNSON: PERCOIDEI
465
with W. D. Anderson). Callanthiids share a flat nasal organ
without laminae, a lateral line that runs along the base of the
dorsal fin, ending near its terminus or continuing along the
dorsolateral margin of the caudal peduncle, and a midlaleral
row of modified scales that bear a series of pits and/or grooves.
The larvae of these two genera appear dissimilar (Fig. 255E, F),
but specimens of Grammatonotus smaller than 1 3 mm are un-
known. Stiginatonolus (based on a small, now lost specimen)
was reported to have three opercular spines, and probably rep-
resents a larval or juvenile anthiine serranid. The family Gram-
matidae, as considered here, contains only Gramma and Li-
pogramma.
Carangidae, Coryphaenidae, Echeneididae, Rachycentridae and
Nematistiidae. — See discussion on utility of larval morphology.
Coracinidae, Drepanidae and Ephippididae. — The family
Ephippididae, as defined here, contains the following genera:
Chaetodipterus. Ephippus. Parapsetttis. Platax. Proteracanthus.
Rhinoprenes and Tripterodon. Ephippidids exhibit considerable
diversity in several features that are more commonly conser-
vative among percoids, such as scale morphology and the struc-
ture and arrangement of median fin supports and predorsal
bones. Nonetheless, monophyly of the family is supported by
shared specializations of the gill arches that include reduction
or absence of the basihyal, absence of the interarcual cartilage,
a relatively large first pharynogobranchial and, most notably, a
peculiar comblike series of large blunt rakers loosely associated
with the anterior margin of the broadened first epibranchial.
Springer (1982; pers. comm.), following some previous authors
(Jordan, 1923; Golvan, 1965) included Parapsettus in the Scor-
pididae. Rhinoprenes was previously treated as a monotypic
family, possibly related to the Scatophagidae (Munro, 1967),
and Proteracanthus as a girellid (Norman, 1966). Although
Drepane may be related to the ephippidids, it does not share
the branchial specializations described above, and lacking fur-
ther evidence of a direct relationship, I treat it separately. Based
on other features of the gill arches a close relationship between
Drepane and Coraciniis seems likely. In both genera the basihyal
is embedded in thick connective tissue and is tightly bound
along the anteroventrally sloping median junction of the hy-
pohyals. In addition, an unusual moveable articulation between
the hypohyals and the anterior ceratohyal allows for dorsoven-
tral rotation of the ceratohyal. Pending further investigation of
the possible affinities of these two genera, I retain them as mono-
typic families. Larval morphology could provide important in-
formation in resolving the relationships among the five ephip-
pidid genera, Drepane and Coraciniis, but to date, only the
larvae of Chaetodipterus have been described (Fig. 256G).
Elassoma— In an extensive comparison of the acoustico-lat-
eralis system of the Centrarchidae, Branson and Moore (1962)
placed the pygmy sunfishes, genus Elassoma. in a separate fam-
ily, based on over 20 "major characteristics." These include
numerous reductions in the laterosensor>' system (e.g., absence
of a lateral-line canal on the body, absence of all infraorbitals
except the lacrimal, absence of the mandibular and angular
lateralis canals, etc.), presence of numerous free neuromasts of
a distinctive form, rudimentary olfactory organ, gill membranes
broadly united across the isthmus, rounded caudal fin, and cy-
cloid scales. To these, I add the following reductive features of
Elassoma, not shared by the Centrarchidae: basisphenoid ab-
sent; endopterygoid absent; ectopterygoid absent or fused to
Table 1 1 9. List of the Families and incertae sedis Genera of the
Suborder Percoidei. * Families with a single genus.
Acanthoclinidae
Acropomatidae
Ambassidae
Aplodactylidae
Apogonidae
Arripididae*
Banjosidae*
Balhyclupeidae*
Bramidae
Caesionidae
Caesioscorpis
Callanlhiidae
Carangidae
Caristiidae*
Cenlracanthidae
Centrarchidae
Cenlrogenysidae*
Centropomidae
Cepolidae
Chaetodontidae
Cheilodactylidae
Chironemidae
Cirrhitidae
Congrogadidae
Coracinidae*
Coryphaenidae*
Datnioides
Dinolestidae*
Dmoperca
Drepanidae*
Echeneididae
Emmelichthyidae
Enoplosidae*
Ephippididae
Epigonidae
Gerreidae
Giganthiidae*
Girellidae
Glaucosomatidae*
Grammatidae
Haemulidae
Hapalogenys
Hemiliiljanus
Howella
Inermiidae
Kuhliidae*
Kyphosidae
Lactariidae*
Lateolahrax
Latrididae
Leiognathidae
Leplobramidae*
Lethrinidae
Lobotidae*
Lutjanidae
Malacanthidae
Menidae*
Microcanthidae
Monodactylidae*
Moronidae
Mullidae
Nandidae
Nematistiidae*
Nemipteridae
Neoscorpis
Opistognathidae
Oplegnathidae*
Ostracoberycidae*
Parascorpididae*
Pempherididae
Pentacerotidae
Percichthyidae
Percidae
Plesiopidae
Pomacanthidae
Pomatomidae*
Polypnon
Priacanthidae
Pseudochromidae
Rachycentridae*
Scatophagidae
Sciaenidae
Scombropidae*
Scorpididae
Serranidae
Sillaginidae
Simperca
Sparidae
Stereolepis
Symphysanodon
Terapondiae
Toxotidae*
palatine; palatine with a single notch-like articulation with eth-
moid cartilage; predorsals usually absent, a single bone present
in some (vs. 3-7 in centrarchids); branchiostegals 5 (vs. 6-7);
principal caudal rays 6-7 -I- 7-8 (vs. 9 + 8); hypurals 1-2 and
3-4-5 fused.
Branson and Moore (1962) concluded that "either the elas-
somids diverged from the centrarchid stock early in the history
of the group or they have entirely different affinities." Subse-
quent classifications (Greenwood et al., 1966; Nelson, 1976)
have continued to treat Elassoma as a subfamily of the Cen-
trarchidae, presumably accepting the conclusion of Eaton (1953,
1956) that Elassoma is a neotenous centrarchid, with most of
its distinctive features having arisen through paedomorphosis.
Weitzman and Fink (1983) attributed similar reductions in the
laterosensory system of small characids to paedomorphosis and
suggested that these characters may be quite labile. These and
other osteological reductions similar to those of Elassoma are
found in other small fishes such as gobioids (Springer, 1983)
and cyprinodontoids (Parenti, 1981), but I know of no such
extreme examples among small percoids.
That the reductive specializations of Elassoma actually rep-
resent character states of earlier developmental stages of cen-
trarchids has never been clearly demonstrated or even ade-
quately investigated, and comparative studies of the osteological
development of these fishes would be necessary to answer this
question. However, a crucial point, that seems to have been
overlooked, is the absence of any other evidence suggesting a
close relationship between Elassoma and the Centrarchidae.
Although I know of no morphological specialization that defines
the family, all centrarchids exhibit a similar mode of nest-build-
466
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
Fig. 254. (A) Symphysanodon sp., 5.1 mm SL; (B) Acropomatidae— Sv/!agropi sp., 8.5 mm SL; (C) \cTopomn\iidie—Acropomajaponicuin,
6.0 mm SL, from Y. Konishi (unpubl.); (D) \cropomai\Aae — Doederleinia herycoides. 8.0 mm SL, from Okiyama (1982b); (E) Polyprion ameri-
canus. 12.2 mm TL, from Sparta (1939a); (F) Slereolepis doederleim. 1.1 mm SL, from Okiyama (1982b); (G) X^ohoxKAat — Lobotessunnamensis.
6.0 mm TL, from Uchida et al. (1958); (H) Hapahgenys sp., 7.3 mm SL, from Okiyama (1982b).
ing and parental-care behavior, and this behavioral "synapo- the search for its origins to the Centrarchidae. Quite the con-
morphy" is not shared by Elassoma (Breder and Rosen, 1966; trary, I believe the affinities of Elassoma will be shown to lie
M. F. Mettee, pers. comm.). Consequently, though Elassoma outside the Percoidei and, perhaps, outside the Perciformes.
may be a product of paedomorphosis, I see no reason to limit My preliminary findings indicate that Elassoma possesses a
i
JOHNSON: PERCOIDEI
467
Fig. 255. (A) Ambassidae— I'elamhassis jacksonensis. 5.5 mm SL; (B) Opislognathidae—Opislognarhus sp.. 6 mm SL; (C) Pseudochromidae,
8.1 mm SL. from Leis and Renins (1983); (D) Acanlhochnidne—Acanlhoclirwslrilmealus. 10.0 mm, from Crossland (1982); (E) Callanthiidae —
Grammatonotus laysanus. 13.7 mm SL, from Leis and Rennis (1983); (F) CaWanXhwdae—Callanlhiaspelontanus. 8 mm TL, from Fage (1918);
(G) Con%xogzA\(i?Le—Congrogadus suhducens. 1 1.8 mm SL; (H) Monodactylidae— A/ono(/acO'/i« sebae. 5.2 mm SL, from Akatsu et al. (1977);
(I) Pempherididae — Pfm/jAmi sp., 5.5 mm SL, from Leis and Rennis (1983); (J) Op\e%mi\\\\dae-Oplcf>nathus fasciatus. 7.5 mm SL.
number of salient features (not mentioned above) that cast doubt
on its affinities with the Percoidei. The second preurai centrum
bears a full neural spine, and there are no autogenous haemal
spines. Strong parapophyses begin on the first centrum, and
pleural ribs may begin on the first, second or third vertebra.
The first neural arch is fused to its respective centrum. The
pelvic fin is inserted well behind the pectoral fin base and the
pelvic girdle docs not contact the cleithra. The first pharyngo-
branchial and interarcual cartilage are absent and what is ap-
parently the uncinate process of the first epibranchial articulates
directly with the second pharyngobranchial. The fourth phar-
yngobranchial, usually cartilaginous in percoids, is absent. The
proximal base of the medial half of the uppermost pectoral ray
does not extend laterally to form a process for articulation with
the scapular condyle (also true of at least some cyprinodontoids
and gobioids). Finally, the ossified portion of the ethmoid con-
sists of two, closely applied, disc-like bones, a condition listed
as one of the defining characteristics of the Atherinomorpha by
Rosen (1964) and Rosen and Parenti (1981). (They did not
discuss the distribution of this character among other groups.
468
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 256. (A) Luljanidae — Luljanus campechanus. 7.3 mm SL, from Collins et al. (1980); (B) Caesionidae— Carao sp. or Gyinnocaesio sp.,
7.8 mm SL, from Leis and Rennis (1983); (C) Leiognathidae— unidentified, 4.8 mm SL; (D) Menidae— A/cne maculala. 4.6 mm SL; (E) Mala-
canlh'idat—Caulolalilus princeps. 6.0 mm SL, from Moser (1981); (F) Ma\acanlh\dae—Hoplolatilusfronlicinclus (head only), 15 mm SL, from
Dooley (1978); (G) Ephippididae—Chaclodiplerusfaber, 9 mm, from Hildebrand and Cable (1938); (H) Pomacanthidae—Centropyge sp., 4.4
mm SL, from Leis and Rennis (1983).
JOHNSON: PERCOIDEI
469
but I have observed a similar condition in the gobiid Dormi-
tator.)
Elassoma seems to exiiibit a confusing mosaic of character
states variously shared with atherinomorphs, cyprinodontoids,
percopsiforms, perciforms, and gobioids. Resolution of the evo-
lutionary affinities of this genus could be important to our un-
derstanding of acanthomorph interrelationships, and I intend
to examine this problem more fully.
Epigonidae. — Eraser ( 1 972a) treated Epigonns. Florencwlla and
Rosenhlattia as a subfamily (Epigoninae) of the Apogonidae,
but I find no evidence to suggest that these genera are closely
related to other apogonids. They are primitive with respect to
apogonids in possessing two pairs of uroneurals and a procurrent
spur (Johnson, 1975), but specialized in several features listed
below. Moreover, the two anal spines of epigonines and apo-
gonids, usually cited as evidence of their close relationship, are
not homologous (see discussion on median fins). The Epigonidae
are here recognized as a distinct family, including Brinkmanella.
Sphyraenops and Eraser's epigonines. These five genera share
the following specializations: rostral cartilage greatly enlarged,
ascending processes of premaxillaries reduced or absent; pre-
maxillary articular cartilages enlarged; endopterygoids large,
metapterygoids notably reduced; infraorbitals more than six.
The larvae of Sphyraenops (Pig. 257A) resemble those of Epigo-
nus (Fig. 257B) but differ in possessing well-developed head
spination.
Girellidae, Kyphosidae, Microcanthidae, Neoscorpis. Parascor-
pididae and Scorpididae. — Springer (1982; pers. comm.), fol-
lowing Jordan (1923) and Golvan (1965), included microcan-
thids, Neoscorpis. Parascorpis and scorpidids in the family
Scorpididae, but no convincing evidence for uniting them has
been presented, and they are treated separately here. The Scor-
pididae is here restricted to Scorpis. Medialuna. Lahracoglossa
and Bathystethus. The latter two genera were treated as a sep-
arate family, Labracoglossidae, by Springer. Scorpidids share
similar meristic and osteological features (not derived) and com-
parable scale morphology. An unusual small slip of muscle ex-
tends from the basioccipital to the first vertebra in Scorpis and
Lahracoglossa. but its presence has not been confirmed in the
other two genera. The larvae of Scorpis and Bathystethus are
undescribed hut those of Lahracoglossa (Fig. 258A) and Me-
dialuna (Fig. 258B) share a similar body form, generalized head
spination, late fin development and pigment pattern with larvae
of the Girellidae (Fig. 258C). Girellids are specialized in several
osteological features with respect to the Scorpididae (see Table
1 20) and have a unique adductor mandibulae in which A, inserts
on the lateral surface of the dentary (Johnson and Fritzsche,
in prep.). The distinctive larval form shared by scorpidids and
girellids suggests that they may be sister groups. Convincing
evidence supporting a close relationship between the Scorpi-
didae and the Microcanthidae (Microcanthus. Atypichthys and
Neatypus) or the Kyphosidae (Kyphosus, Seclator and Her-
nwsilla) is lacking. Furthermore, the larvae of the latter two
families (Figs. 259G, J) do not possess the salient features of
scorpidid or girellid larvae, but more closely resemble those of
the Teraponidae (Fig. 259H). The larvae of Neoscorpis and Par-
ascorpis are unknown, and available anatomical information is
insufficient to clarify the systematic position of these two genera.
Malacanthidae. — See discussion on utility of larval morphology.
Moronidae {Morone and Dicentrarchus), Lateolahrax and 5/>j-
/perca.— Gosline (1966) included the Moronidae (using the name
Roccus). Lateolahrax and Siniperca (=Coreoperca) in his "es-
tuarine and freshwater percichthyids." I treat these separately,
because I lack evidence of their affinities with the Fercichthyi-
dae, with one another, or with any other percoid group. It is
interesting to note that the Moronidae share one of the two
synapomophies of the Centropomidae described by Greenwood
(1976)— the lateral line extends almost to the posterior margin
of the caudal fin. In addition, moronids have an auxilliary row
of lateral line scales on the caudal fin above and below the main
row, as does the centropomid Lates. Although both of these
conditions occur elsewhere in generalized percoids (e.g., Neo-
scorpis. some species of Lutjanus. and the percid subfamily
Luciopercinae) and may actually be primitive for the Percoidei
(Springer, 1983), the possibility of a moronid-centropomid
relationship seems plausible and should probably be investi-
gated further. Unfortunately, as is typical of most fresh or brack-
ish water spawners, the larvae of these groups (Fig. 260) exhibit
relatively direct development and consequently offer little phy-
logenetic information.
Percichthyidae. — The Percichthyidae of Gosline (1966) repre-
sents a polyphyletic assemblage defined on the basis of shared
primitive features. I am unable to find synapomorphies that
support recognition of the assemblage as a monophyletic group.
I restrict the Percichthyidae to the following genera, which occur
only in freshwaters of Australia and South America: Percolates
(brackish water), Plectroplites. Macquaria. Maccullochella. Per-
cichthys, Percilia. Bostockia. Gadopsis. Nannoperca. Edelia. and
Nannatherina. The monophyly of the family is supported by a
series of nested synapomorphies, only a few of which are men-
tioned here. The scales of most of these genera are similar and
unlike those of the excluded genera in having the posterior field
filled with simple, only slightly amputated (see McCully, 1970),
needle-like ctenii (those of Bostockia. Gadopsis and Nannath-
erina are secondarily cycloid). The three most generalized gen-
era. Percolates, Plectroplites. and Macquaria are very similar
biochemically [MacDonald (1978) synonymized them on this
basis], and the latter two share two morphological specializa-
tions with Macidlochella. Percichthys. Percilia. Bostockia and
Gadopsis: enlarged sensory pores on the dentary and a separate
inner division of adductor mandibulae section A,. The three
most derived genera, Nannoperca. Edelia and Nannatherina
(heretofore treated as kuhliids) share with Bostockia a similar
vertebral number (29-33), a distinctive asymmetrical nasal ro-
sette, and a number of reductive specializations (absences of the
subocular shelf, procurrent spur, and supracleithral sensory ca-
nal, reduced numbers of procurrent caudal rays, dorsal spines,
branchiostegals and trisegmental pterygiophores, and an inter-
rupted or absent lateral line). Systematic placement of the enig-
matic Gadopsis has proved problematic, even in recent years.
It has generally been treated as a monotypic family and variously
assigned to the Percoidei (Greenwood et al., 1 966), Ophidioidei
(Gosline, 1 968), Perciformes with proposed affinities to the Tra-
chinoidei and Blennioidei (Rosen and Patterson, 1969) or a
separate order Gadopsiformes (Scott, 1962). The percoid affin-
ities of Gadopsis are manifest in the anatomy of the dorsal gill
arches, caudal skeleton and median fin supports. Its affinities
with the Percichthyidae are indicated by a number of features
shared with some percichthyid genera, including the configu-
ration of the adductor mandibulae noted above. Gadopsis shares
470
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 257. (A) Epigonidae—Sphyraenops bairdianus. 6.8 mm SL; (B) EpxgomAae—Epigonus sp., 14.0 mm SL; (C) Howella sp., 6.0 mm SL;
(D) \po%.on\dne — Pseudamia sp. or Pseudamiops sp., 8.7 mm SL, from Leis and Rennis (1983); (E) Apogonidae — foa brachvgrainma. 4.2 mm
SL, from Miller et al. (1979); (F) Apogonidae — unidentified, 4.2 mm SL, from Leis and Rennis (1983); (G) Apogonidae— unidentified, 5.0 mm
SL, from Leis and Rennis (1983); (H) Sciaenidae— 5rW///er tanceotalus. 6.2 mm SL, from Powles (1980).
the asymmetrical nasal ro^eXXe oi Bostockia. Nannoperca. Edelia
and Nannalherina and all reductive specializations of those gen-
era noted above, except the reduced lateral line and branchio-
stegal number. Specializations shared with Bostockia alone in-
clude a tubular anterior nostril placed near the margin of the
lip and absences of the basisphenoid, medial tabular, and third
epural. Based on this evidence, Gadopsis appears to be most
closely related to Bostockia, however it bears a strong superficial
resemblance to Macullochella and shares the premaxillary fre-
num of that genus.
Adult Morphology
The scope of morphological diversity exhibited within the
Percoidei surpasses that of all other perciform suborders. Al-
though many percoids have a generalized bass-like or perch-
like physiognomy, extremes of adult body form range from deep
bodied, compressed or "slabsided" fishes, such as the ephip-
pidids, chaetodontids and menids to extremely elongate forms
like the cepolids and the eel-like congrogadids. Add to this the
exceptional variability in fin conformation, ornamentation of
head bones, squamation. jaw configuration, and internal osteo-
logical features, and the suborder Percoidei presents an im-
pressive heterogeneous array of forms. Lack of progress in elu-
cidating percoid phylogeny is largely attributable to this
somewhat overwhelming diversity and the ostensible conver-
gence (particularly in reductive traits) that seems to have char-
acterized percoid evolution. To date, no familial phylogeny,
cladistic or otherwise, has been proposed for the suborder. The
limits and monophyly of many of the component families are
not clearly defined and the affinities of numerous genera remain
unresolved. Superficial knowledge of basic percoid anatomy and
an inadequate understanding of character distribution and vari-
JOHNSON: PERCOIDEI
471
Fig. 258. (A) Scorpididae— Labracoglossa argenliventris, 9.9 ininSL;{S)ScoTpididat—Medialunacaliforniensis, 10.1 mm SL;(C)Girellidae—
Girella nigricans. 10.9 mm SL; (D) Leptobramidae — Z,fp/o/)rama mulleri. 7.2 mm SL; (E) CheWodacXyWdm—Palunolepishrachydactylus. 8.3 mm
SL; (F) C\Tr\\\\\dae—Amhlycirrhituspinos. 13.2 mm SL; (G) PoTm\om\dae— Pomatomus satlalrix. 7.3 mm TL. from Pearson (1941); (H) Nem-
ipteridae— unidentified, 5.1 mm SL, from Leis and Rennis (1983); (I) Spandae—Acanthopagrus cuvieri. 8 mm SL, from Hussain et al. (1981); (J)
Cenlracanlhidae— Plerosmaris axillaris, 7.7 mm SL, from Brownell (1979).
ability, basic to cladistic outgroup comparison, have seemingly
inhibited, or at least hindered, meaningful comparative studies
within the Percoidei.
Because the group is so large, these problems will necessarily
continue to plague studies of percoid relationships. Outgroup
comparisons based on a single family are speculative without
evidence for a sister group relationship, and broader surveys of
each character are frequently impractical if not impossible. One
approach that can gradually alleviate this problem is the cu-
mulative tabulation of characters and character states. Com-
parative tables document the distribution of morphological fea-
tures throughout the suborder and the variability of these features
within families, and they accordingly offer the most complete
foundation for outgroup comparison. Furthermore, they pro-
vide information about the plasticity of various complexes, al-
low identification of characters most frequently subject to con-
472
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
JOHNSON: PERCOIDEI
473
.%
^-'^V,-:-.
Fig. 260. (A) Centrarchidae— .•)«iWop//fM rupestris. 10.5 mm TL, from Fish (1932); (B) Simperca (=Coreoperca) kawamebari. 9.0 mm TL,
from Imai and Nakahara (1957); (C) Pemdae — Perca flavescens. 14.2 mm TL, from Mansueti (1964); (D) Percichthyidae— A/aa-w//oc/!e//a
macquanensis, size unknown, from Dakin and Kesteven (1938); (E) Lateolabrax japomcus. 13.7 mm TL, from Mito (1957b); (F) Moronidae —
Morone amencana. 13.2 mm TL, from Mansueti (1964); (G) Centropomidae— Ce«rrapo«ii« undecimatis. 6.3 mm SL, from Lau and Shafland
(1982).
vergence and convincingly document the uniqueness of derived
features. With this in mind, I have compiled information about
selected morphological features of adults (Table 1 20) and larvae
(Table 121) for each percoid family or inceriae sedis genus. This
information was compiled from the literature (particularly the
meristic data) and from my own examination of cleared and
stained specimens and radiographs. Data for a few groups were
compiled by experts working on those groups. For many fam-
ilies. I examined at least one representative of each genus, but
obviously this was not always possible and only in a few of the
smaller families were all species examined. As a consequence,
this data will not reflect the full range of variability for every
family but should represent a reasonably close approximation.
Most features considered in Table 1 20 are discussed below.
Fins— The primitive perciform complement of one spine and
five rays (I, 5) in the pelvic fin is the most consistent feature of
percoid fins. A single spine is always present and fewer than
five soft rays are found only in the Acanthoclinidae (I. 2), Con-
grogadidae (I, 2-4 or absent), Plesiopidae (I, 4), Pseudochro-
midae (I, 3-5) and the percichthyid Gadopsis (I, 1).
The primitive and most common number of principal caudal
fin rays (branched rays + 2) is 9 + 8. Where reductions occur
(in 18 families) they usually involve one fewer principal ray
dorsally and/or ventrally and are frequently consistent within
families, e.g.. 8 + 7 in Cheilodactylidae. Chironemidae. Cir-
rhitidae. Latrididae and Mullidae. and 8 + 8 in Acanthoclini-
dae, Priacanlhidae, and Scatophagidae. The most extreme re-
duction (4-6 branched + 4-8 branched) is seen in the
Congrogadidae. The only apparent increases, 10 + 9 found in
some grammatids and plesiopids, do not result from an in-
creased number of rays articulating with the hypurals, but from
branching of the outermost hypural-associated rays. Numbers
of procurrent or secondary caudal rays dorsally and ventrally
Fig. 259. (A) Ge:m\(ia.e— Eucinostomus sp., 8.7 mm SL; (B) HaemuWAae — Xenistius californiensis. 6.5 mm SL; (C) HaemuXiAae — Pseudo-
prislipoma nigra. 5.8 mm SL, from Leis and Rennis (1983); (D) HaemuVxdae-Conodon nobdis. 9.8 mm SL; (E) Mullidae. 8.2 mm SL. from
Miller et al. (1979); (F) Sillaginidae— .S'///tig<) .«/;ama. 9.0 mm TL, from llchida et al. (1958); (G) M\!:TocaM\\'\dae — Microcanlhussthgatus. 7.1
mm TL. from Uchida et al. (1958); (H) Tcraponidae— F/icrapo/i iheraps. 9.5 mm. from Zvjagina (1965b); (I) Emmelichthyidae— £o'''"'octe
schlegeh. 6.9 mm TL. from Nakahara (1962); (J) Kyp\\o%\dae- Kyphosus anerascens. 9.8 mm TL. from Uchida et al. (1958).
474
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 261. {A)Coryphaemd!ie—Coryphaenahippurus. 8.5 mm SL; (B) Rachycenlridae—Rachycenlroncanadum. 9.0 mm SL; (C)Echeneididae-
Echeneis sp., 8.8 mm SL; (D) Caristiidae— Canif/i« sp., 10.1 mm SL; (E) Bramidae— firaAjja dussuimeri, 6.5 mm SL, from Mead (1972).
range from 0 in the Congrogadidae to 19 in the Sillaginidae, the
most common numbers being 8-14.
One of the most variable aspects of percoid physiognomy is
the form and composition of the dorsal fin. Even the most
consistent feature, the presence of spines, does not characterize
all percoids. Absence of dorsal spines in six percoid families
appears to have originated by at least two different mechanisms.
In Bathyclupea, it is obvious that the spines have been lost
because the spinous pterygiophores are still present and the soft
rays occupy a position posterior to them. In Coryphaena, how-
ever, Potthoff (1980) showed that although the anteriormost 3-
4 pterygiophores bear soft rays, they are of the type that normally
support spines. This suggests that the absence of spines in Cory-
phaena is the result of transformation, rather than loss, of pre-
existing elements. Absence of spines in the Bramidae, Caristi-
idae, some cepolids and some congrogadids is also probably the
result of transformation.
Spines are present anteriorly in the dorsal fin of all other
percoids, ranging from 1 in some malacanthids and pseudo-
chromids to XXI in some acanthoclinids. Dorsal soft rays range
Fig. 262. (A) Chaetodontidae— unidentified, 10 mm, from Burgess (1978); (B) C\ae\oAon\\Aae—Forcipigerlongiroslris. 17 mm TL, from
Kendall and Goldsborough (191 1); (C) Chaetodontidae— C/jWotom sp. or Coradion sp., 6.5 mm SL, from Leis and Rennis (1983); (D) Scato-
phagidae— 5ca/o/7AagMi argiis, 10 mm SL from Weber and de Beaufort (1936); (E) Scombropidae — 5fom/)TOpi hoops. 6.2 mm SL, from Uchida
et al. (1958); (F) Lethrinidae— Z.e?/!n>ii/i nematacanlhus, 6.1 mm SL, from K. Mori (unpubl.); (G) CepoMiae—Acanthocepola sp., 9.7 mm SL,
from Okiyama (1982b); (H) Priacanlhidae — unidentified. 4.6 mm SL. from Leis and Rennis (1983); (I) Priacanthidae— /'nacanr/iMisp., 10.9 mm
SL, from Leis and Rennis (1983); (J) PenXace^toUd&c — Pseudopentaceros richardsoni. 15 mm SL.
JOHNSON: PERCOIDEI
475
476
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 120. Selected Morphological Features of Adult Percoidel Abbreviations and definitions: SS— supernumerary (non-serial) spines
(or soft rays) on first anal pterygiophore (see Johnson, 1980); D— dorsal fin; A— anal fin; Tnseg. pteryg. — pterygiophores with proximal, medial
and distal radials separate; Stay — separate bony element posterior to ultimate pterygiophore in D and A; Predorsal formulae — based on Ahlstrom
et al, (1976); P— pterygiophore with no supernumerary spines or soft rays; H — hypurals; E— epurals; U — uroneurals; Ah— autogenous haemal
spines; pH — parhypural; UR — urostyle; Proc spur— procurrent spur (see Johnson, 1975); PU3 cart- radial cartilage anterior to neural and haemal
spines of third preural centrum; BR — branchiostegals; lAC— interarcual cartilage; Cy— cycloid; Ct— ctenoid, ctenii free from posterior margin;
Ct— ctenoid, ctenii continuous with posterior margin; and br— branched caudal fin rays. With the exception of (SS), parentheses enclose features
known to characterize only some members of a group.
Vertebrae
Dorsal fin
Tbiseg.
PTERYG-
D
— Slav
A
Pelvic
fin
Predorsal formulae
CAUDAL RN
Pnncipal
Anal fin (SSl
Procurrent
Acanthoclinidae
10 +
11 +
18
17
XVIII-XXI, 3-4
VIII-XI,4(l-2)
h
1,2
0/0/1/1 + 1/
8 + 8
3-4 + 3-4
11 +
18
12 +
18
13 +
16
Acropomatidae
10 +
15
VII-X-I, 8-10
or
'■a*
1,5
/0+0/0 + 2/1 + I/
0/0/0 + 2/1/
9 + 8
9-13 + 9-13
IX-X, 10
0/0/0 + 2/1 + 1/
II-III, 7-9 (2)
Ambassidae
10 +
14
Vll-I, 8-11
111,7-10(2)
0
0
1,5
0/0/0+1/1/
0/0/1 + 1/1/
9 + 8
7-11 + 7-10
Aplodactylidae
15 +
16 +
20
19
XIV-XXI, 16-21
III, 6-9 (2)
I*
1,5
0/0 + 0/2+1/1/
0/0 + 0/2/1 + 1/
8-9 + 7
12-14+12-13
Apogonidae
10 +
10 +
14
15
VI-VIII-I, 8-14
11,8-18(1)
m*
1,5
various:
0/0/0 + 2 or 1/
to
///2 or 1/
9 + 8
6-10 + 5-10
Arripididae
10 +
15
IX, 13-19
\*
I, 5
0/0/0 + 2/1 + 1/
9 + 8
111,9-10(2)
7 + 6
Banjosidae
11 +
14
X, 12
9
1,5
0/0 + 2/1/1/
9 + 8
III, 7
6 + 5
Bathyclupeidae
9 + 22
10 + 21
9
i-
1,5
0/0/0//P+ P/P/P+ P/P/P/P/ 1 /
9 + 8
I, 26-27 (0)
9 + 8
Bramidae
Braminae
T;36-
-47
30-38
I*
1,5
0/0/0/P/P/l + l/
0/0/0/P/l/
9 + 8
21-30(2-3)
7-8 + 7-8
Pteraclinae
T;45-
-54
46-57
39-50 (?)
'-a-
1,5
l + l + l + l,etc./l/l/
9 + 8
5 + 5
Caesionidae
10 +
14
IX-XV, 9-21
111,9-13(2)
w
1.5
0/0/0 + 2/1 + 1/
/0 + 0/2/I + 1/
9 + 8
7-10 + 5-10
Caesioscorpis
10 +
15
XI, 20-21
III, 18-20(2)
6-8
6-8
1,5
0/0 + 0/2/1
9 + 8
9-11+8-10
Callanthiidae
10 +
14
X-XI, 8-12
'v
1,5
0/0/2/1 + 1/
8-9 + 7-8
111,9-12(2)
5-9 + 5-9
Carangidae
10 +
10 +
14
15
IV-VIII-I, 17-44
I-II-I, 15-39(1-2)
0-3
0-2
1,5
see section on
Carangidae
9 + 8
8-14 + 8-12
10 +
16
10 +
17
11 +
13
11 +
14
Caristiidae
T: 35
-40
32-40
18-21 (1)
O9
0 ■
1,5
l + 1 + I + l + l + I/l/I/
l + I + I + I/I/I/
9 + 8
7-8 + 7-8
Centracanthidae
10 +
14
XI-XIII, 9-11
111,9(2)
'-'* +
1-4
1,5
0/0 + 0/2+1/1/
0/0/0 + 2/1/
9 + 8
9-10 + 8-10
Centrarchidae
11-
+
4
V-XIII, 9-16
3-6
3-6^
1,5
3-7 predorsals
1-3 sup. spines
9 + 8
II-VIII, 8-19 (2-3)
5-10+5-9
15-
9
T:28
-32
Centrogenysidae
11 +
14
XIIl-XIV, 9-11
III, 5 (2)
0
0 +
I, 5
0/0 + 2/1/1/
7 + 7
5 + 5
JOHNSON: PERCOIDEI
477
Table 120. Extended.
SKELETON
H/E/U/Ah
H Fusions
Proc
spur
"PU3
can.
BR
lAC
Scales
3/3/0/0
+
6
+
Ct
pH-l-2;3-4-UR
or
Cy
5/3/2/2
+ +
+
7
+
Ct
or
Cy
3/2/1/1
1-2: 3-4
+
6
+
Cy
5/3/1/2
?
6
7
Cy
2-5/2-3/0-1/0-2
various
+
7
+
Ct
or
Cy
5/3/2/2
+ -
+
7
+
Ct
5/3/2/2
+ +
7
+
Ct
5/3/2/2
+ +
+
7
+
Cy
5/3/2/2
+ -
+
7
+
Ci'
5/3/2/2
+ -
+
7-8
+
Cf
3 or 5/3/2/2
(1-2; 3-4)
7
7
+
Ct
5/3/2/2
+ +
?
, 7
+
Ct
3/2-3/1/1
1-2; 3-4
+
6
+
Ct
2-3/2-3/1-2/2
1-2; 3-4-(5)
7-9
+
Cy
5/3/?/2
3/3/2/2
1-2; 3-4
5/3/1-2/2
7
+
Cy
6
+
Ct
6-7
+
Ct
from as few as 3 in some acanthoclinids to 89 in some cepolids.
Within families, the range of dorsal fin ray counts may be rel-
atively restncted as in the Lutjanidae (X-XII, 10-17) or quite
broad, as in the Sciaenidae (VII-XV — 1, 17-46).
In most percoids the spinous and soft portions of the dorsal
fin are continuous, but gradual shortening of the posteriormost
spines results in a variously developed cleft or apparent sepa-
ration. Where this cleft is present, the ultimate spine is notably
longer than the penultimate and is considered to form the first
element of the soft dorsal portion of the fin. Some groups (e.g.,
Pseudochromidae, Grammatidae, Plesiopidae, Priacanthidae,
Cepolidae) do not develop this cleft. Others, such as the Ac-
ropomatidae, Ambassidae, Apogonidae, Emmelichthyidae, En-
oplosidae and Epigonidae have such a well-developed cleft that
the spinous and soft portions of the fin appear completely sep-
arate. Pterygiophores usually continue beneath the resultant gap
and may even bear minute spines. The Mullidae and Echenei-
didae are exceptional in having no pterygiophores below this
gap. Extreme separation of the spinous and soft dorsal fins oc-
curs only in the Echeneididae, where the spinous dorsal fin has
been modified as an attachment disc and has moved far forward
to cover the flattened cranium.
The anal fin of percoids is less variable in form and com-
position than the dorsal fin. The most common, and apparently
primitive condition is three anal spines. The first anal pteryg-
iophore is larger than the succeeding pterygiophore and bears
the first two spines in supernumerary (non-serial) association
and the third in serial association (see Johnson, 1980). Scato-
phagids and some chaetodontids and pomacanthids have four
spines, the first two supernumerary. Centrarchids have from
two to eight spines, pentacerotids from two to six and gerreids
from three to five, all with the first two supernumerary. The
only other percoids characterized by more than three anal spines
(eight to eleven) are the Acanthoclinidae, where one or two may
be supernumerary. Several groups have fewer than three anal
spines, and, as in the dorsal fin, it is important to understand
the nature of this reduction. Apogonids, for example, have only
two spines and only one of these is supernumerary, suggesting
that the anteriormost spine was lost. The mesopelagic epigon-
ines (sensii Eraser, 1 972a) have usually been treated as a subfam-
ily of the Apogonidae, for they also have only two anal spines.
The epigonine anal spines, however, are both supernumerary
(as are those of the Sciaenidae), suggesting that the usually spi-
nous third (serial) element has not transformed into a spine.
Hence, the two anal spine conditions of epigonines and apo-
gonids are not homologous. In bathyclupeids, the single anal
spine is serially associated with the first pterygiophore, sug-
gesting that the first two spines have been lost. Only a few groups,
Bramidae. Caristiidae, Congrogadidae, Coryphaenidae and some
cepolids and grammistine serranids, lack anal spines. The pres-
ence of 1-3 supernumerary elements on the first pterygiophore
in all these groups indicates transformation rather than loss of
the pre-existing spines. Anal soft rays range in number from 4
in the Acanthoclinidae to 101 in the Cepolidae and, with some
exception, the range of variability within families is comparable
(frequently within two or three rays) to thai of the dorsal soft
rays.
4/3/1/2
1-2
Ct
Predorsal bones. — In most percoids, one to three strut-like bones
precede the anteriormost pterygiophores of the dorsal fin. It has
been proposed (Smith and Bailey, 1961), but never conclusively
demonstrated, that these predorsal bones were derived from
true pterygiophores that once bore spines or rays, but Eraser
478
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 120. Continued.
Dorsal fin
Anal fin (SS)
Triseg.
PTERYG.
D
-Slay
A
Predorsal formulae
CAUDAL RN
Pnncipal
fin
Procurrent
1,5
0/0/0 + 2/1 + 1/
0/0/0+1/1 + 1/
9 + 8
8-12 + 7-11
Cenlropomidae
10 + 14
11 + 14
Cepolidae
Cepolinae
1 2 + 44-66
14 + 55
16 + 53
Owstoniinae
11 + 17
13 + 16
14 + 16
Chaetodontidae
10 + 14
11 + 13
Cheilodactylidae
14 + 21
Chironemidae
13 + 20
14 + 19
Cirrhitidae
10 + 16
Congrogadidae
12-19
+
34-64
Coracinidae
Coryphaenidae
Datnioides
Dinolestidae
Dmoperca
Drepanidae
Echeneididae
Emmelichthyidae
Enoplosidae
Ephippididae
Epigonidae
Gerreidae
10 + 15
13-15
+
17-19
T: 30-34
10 + 14
10 + 15
11 + 14
10 + 14
VII-VIII. 8-13
III, 6-9 (2)
O-lII, 65-89
0-1,62-101
III-IV, 21-27
I-II, 14-19
Vl-XVl. 15-30
III-V. 14-23(2)
XIV-XXII. 15-39
111,6-19(2)
XlV-XVl, 15-21
III, 6-8 (2)
X, 11-17
III, 5-7(2)
O-II. 33-76
28-63
X. 18-23
III, 13-14(2)
52-66
23-31 (2)
XII, 15
III, 15-28(2)
VII-VlII-1. 7-10
I-III, 7-9(1-2)
IX-X. 9-17
III-V, 13-17(2)
0-5
0-4
111.9(2)
1
10 + 17
Vliii-I, 18-19
1,26-27(1)
0-1
0-1
10 + 16
XI, 17-19
III, 11-13(2)
4
3
10 + 14
XIII-IX, 19-22
111, 17-19(2)
0
0
12-18
+
14-22
T:26-40
IX-XXVIII- 17-42
II, 14-36(1)
0
0
10 + 14
XI-XIV, 9-12
III, 9-1 1 (2)
6-8
6-8
10 + 16
VIII-I, 14-15
III, 14-15(2)
1-7
4-8
10 + 14
V-IX, 18-40
0(11*)
0
0^
0-( + )
0
0 +
0
0 +
0
0 +
0 _
0
1*
0 _
0
5
1
J^
1
4
3
0( + )
+
+
0(11*)
*Ephippus
0-1
( + )
0-1
0-2
0-2
I, 5
1,5
I, 5
I, 5
I, 5
I, 5
1,5
mm
HUM
0/0 + 2/1/
0/2/1/1/
0 + 0/2+1 + 1/1/
0 + 0/2+1/1 + 1/
0/0/2+1/1/
0/0+0/2/1 + 1/
0/0/0 + 2/1 + 1/
I, 2-4 0/0/0/P+ 1
or 0/0//P+1
absent 0/0//P + P
///P+P/
1,5 0/0/0 + 2/1 + 1/
l-3+l + l + l,etc./l/I/
5
0/0/0 + 2/1/
5
0/0/0+1/1 + 1/
0/0/0/1 + 1/
5
0/0/0 + 2+1/1/
5
0/0 + 0/2/1 + 1/
5
absent; D,
on head
5
0/0/0 + 2/1/
5
0/0/0 + 2/1 + 1/
5
0/0 + 0/2/1 + 1/
0+0 + 0//2/1 + 1/
0 + 0/2/1 + 1/
0/0/0+1/1/
0 + 0/0/P/l
5
0/0/0 + 2/1 + 1/
/0+0/0 + 2/1 + 1/
/O + 0/O+l/l/
5
0/0/0 + 2/1 + 1/
6-7 + 6
1-2+1-2
8 + 7
3-4 + 3-4
9 + 8
2-4 + 2-3
8 + 7
9-1 1 +8-10
8 + 7
13-16+10-12
8 + 7
9-14+10-13
4-6 br + 4-8 br
0-4+0-3
9 + 8
9 + 8
9 + 8
10-
-14+10-
9 + 8
6 + 7
9 + 8
11 + 11
9 + 8
13+12
9 + 8
5+4
9 + 8
14
8-13 + 7-13
9 + 8
7-8 + 7-8
9 + 8
5-6 + 5-6
9 + 8
3-7 + 3-6
9 + 8
9-10 + 7-10
9 + 8
9-11+9-10
JOHNSON: PERCOIDEI
479
Table 120. Continued. Extended.
( AUnAl
SMIEION
H E 11 Ah
H Fusions
Proc
spur
Wi
cart.
5/2-3/1-2/2
+ +
—
+
2-4/1-2/1/2?
__
(1-2; 3-4; 5 absent)
7
3-4/3/1/2
_ _
(1-2; 3-4)
-
5/3/2/2
5/2-3/1/2
5/3/1/2
_ _
—
?
5/3/2/2
—
+
2/0-2/0/0-1
pH-1-2; 3-4-UR
5/3/2/2
3/1/1/2
1-2; 3-4
+ +
Ct
6
+
Cy
6
+
Cy
6
+
Ct
3-6
-
Cy
6
-
Cy
6
r
or
Cy
Cy
Ct
Cy
5/3/2/2
+ +
+
6
+
Ct
-
5/3/2/2
+ +
?
7
+
Cy
5/3/2/2
+ +
7
+
Ct
5/2-3/2/2
+ +
+
6
+
Cy
5/2/1/1-2/
- +(-)
^ 8-11
-
Cy
5/3/2/2
+ +
7
r
Ct
—
?
or
5/3/2/2
+ +
?
7
+
Cy
4-5/3/2/2
+ +(- -)*
6
—
Ct
(2-3)
( + )
*Rhinoprenes
or
Cy
5/3/2/2
+ +
7
+
Ct
+
or
3 or 5/3/2/2
(1-2; 3-4)
+ +
+
6
+
Ct
( 1 972a) argued that the first three predorsal elements of percoids
may represent supraneurals. Ahlstrom et al. (1976) recognized
the importance and utility of considering patterns of predorsal
bones in early life history studies, and further developmental
studies could resolve the origin of these elements.
The most common and presumably primitive number of pre-
dorsal bones in percoids is three; Table 120 shows that over
half of 91 percoid groups (families and incertae sedis genera)
have three predorsal bones exclusively, with three predorsals
occurring in at least some members of 66 groups. The first dorsal
pterygiophore inserts in the third intemeural space in at least
some members of 69 groups, bears two supernumerary spines
in some members of 69 groups and exhibits both conditions in
57 groups. Therefore, the most common and ostensibly prim-
itive predorsal formulae (using that defined by Ahlstrom et al.,
1976) for the Percoidei are 0/0/0 + 2/ and 0/0 + 0/2/. The 0/0/1
pattern, considered by Smith and Bailey (1961) to be primitive
for percoids occurs in only six families, frequently in the more
derived members. Furthermore, Fraser (1972a) noted that der-
ivation of the 0/0/0 + 2/ or 0/0 + 0/2/ patterns from the 0/0/1
pattern by backward shift of the first dorsal spine, hypothesized
by Smith and Bailey (1961), is untenable and inconsistent with
pterygiophore interdigitation. On the other hand, the 0/0/0 + 2/
pattern could be easily derived by a posterior shift of the first
dorsal spine in the 0/0/1 + 1/ pattern that characterizes many
beryciforms, including holocentrids and diretmids. This latter
pattern is found among percoids only in some ambassids.
Departures from the primitive predorsal pattern have appar-
ently arisen independently in many families. In anterior shifts
of the dorsal fin origin a compound first pterygiophore with two
supernumerary spines is frequently retained, but it is invariably
absent in posterior shifts. A possible conclusion is that forward
shifts result from anterior displacement of the pterygiophores,
whereas posterior shifts result only from loss of spines. Reduc-
tions in numbers of predorsal bones to fewer than three are
almost certainly the result of simple losses as opposed to trans-
formations, even when these reductions are accompanied, as
they are occasionally (e.g., Chaetodontidae, Scatophagidae, Pen-
tacerotidae, Priacanthidae), by a forward shift of the dorsal fin
origin. Transformations of pre-existing predorsal bones to spi-
nous pterygiophores would require the unlikely addition of de
novo spines and distal radials, and modification of musculature.
More than three "predorsal bones" are found in at least some
members of 13 percoid groups, with a corresponding posterior
shift of the dorsal fin origin. The additional elements are usually
distinguishable from the anterior three ("true") predorsals. In
Bathyclupea, Braminae, some carangids, Congrogadidae, Mene,
Neoscorpis. Platax. some pempheridids and Toxotidae, these
additional spineless elements (designated P in Table 120) re-
semble pterygiophores, may have separate distal elements, and
often articulate with succeeding similar elements or with the
anterior-most spine-bearing pterygiophore. In Bathyclupea and
Toxotidae. they are also separated from the true predorsals by
one intemeural space. In the remaining groups with more than
three "predorsals" (some percichthyids and centrarchids, Brink-
manella and Leptohrama). the additional elements are not mor-
phologically distinguishable from the anterior three, but, as in
the other groups, the dorsal fin originates posterior to the third
intemeural space (except in Brinkinanella), and it seems likely
that these elements were also derived from pre-existing pteryg-
iophores. Studies of the sequence of development of predorsal
480
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 120. Continued.
Anal lit! (SS)
Triseg.
PTERYG,
D
-Slav
A
Pelvic
fin
Predorsal formulae
CAL'DAL RN
Pnncipal
Procurrenl
9 + 8
10 + 9
9 + 8
9-14 + 8-12
9 + 8
7-8 + 7-8
8-10 + 7-9
2-8 + 2-7
9 + 8
9-14 + 8-13
9 + 8
6 + 5-6
9 + 8
11 + 10
9 + 8
9-10 + 9-10
9 + 8
10-12+10-12
9 + 8
11-13+10-12
9 + 8
Giganthiidae
10 + 15
Girellidae
11 + 16
11 + 17
14 + 20
Glaucosomatidae
10 + 15
Grammatidae
10 + 15
10 + 17
10 + 18
Haemulidae
10+ 16
11 + 16
Hapalogenys
10 + 14
Hemilutjanus
10 + 15
Howella
10 + 16
Inermiidae
12 + 14
13 + 13
Kuhliidae
10 + 15
Kyphosidae
10 + 15
10 + 16
Lactariidae
10 + 14
Laleolabrax
17 + 18
Latrididae
14 + 21
Leiognathidae
10 + 14
Leptobramidae
10 + 14
Lethrinidae
10 + 14
Lobotidae
11 + 13
12 + 12
Lutjanidae
10 + 14
Malacanthidae
10 + 14
11 + 14
11 + 16
Menidae
10 + 14
Microcanlhidae
10 + 15
Monodactvlidae
10 + 14
10 + 15
Moronidae
11 + 14
12 + 13
IX. 13
111,8(2)
XII-XVl. 11-17
III, 10-13(2)
VIII, 1 1
111,9(2)
XII-XIV. 8-10
III, 7-11 (2)
IX-XIV, 1 1-26
111,6-13(2)
XI-XIV, 14-19
111,9-13(2)
X, 10-11
III, 9 (2)
VIII-I, 9
111.7(2)
X-IIorXVlI, 9-
10
111,8-10(2)
X. 9-12
111,9-13(2)
X-XII. 11-15
III, 10-16(2)
VlI-VIII-1. 20-22
III, 25-28 (2)
XII-XIV, 12-16
111,7-10(2)
XIV-XXIII, 23-40
III, 18-35(2)
VIII, 15-16
III, 14(2)
IV, 16-18
111.26-30(2)
X. 9-1 1
111,8-10(2)
XII, 15-16
III. 11 (2)
X-XII, 10-17
111,7-11 (2)
I-X, 14-60
I-II, 11-55(1-2)
IV, 38-42
III, 28-32 (2)
X-XI. 16-22
111, 13-19(2)
VII-IX. 26-36
III, 27-37 (2)
VIll-X-1, 10-13
111,9-12(2)
6
0-1
0-1
7
0
0
0
0
6
6
0
0
0
0
4-7
5-8
1-7
3-5
3-5
2-3
6
3
0
0
15-16
13
0
0 "
2-3
2-3 '
0
0 "
1-7
1-7 "
?-l7
?-14 "
0
O"
10-16
10-15 '
0-2
0-1 "
2-4
2-4 "
1.5
0/0/2/1 + 1/
1. 5
0/0 + 0/2/1 + 1/
0/0/2/1 + 1/
I. 5
0/0/0 + 2/1 + 1/
1,5
0/0/0 + 2/1 + 1/
0/0/0+1/1 + 1/
I, 5
0/0/0 + 2/1/
0/0 + 0/2+1/1/
0/0 + 0/2/1/
1,5
0/0 + 0/2/1 + 1/
1,5
0/0/0 + 2/1 + 1/
1,5
0/0/0 + 2/1 + 1/
1,5
/0/0 + 2/1/
/0 + 0/0 + 2/1/
1,5
0/0/0 + 2/1 + 1/
0/0/0 + 2/1/
1,5
0/0/0 + 2/1/
1,5
0/0/0 + 2/1 + 1/
1,5
0/0/0+1/1 + 1/
1,5
0/0 + 2/1 + 1/1/
0 + 0/2/1 + 1/1/
1,5
0/1/1/1/
I, 5
0/0/0/0//P/P/P/P/P/ 1 / 1 + 1 + 1 /
I, 5
0/0 + 0/2+1/1/
1,5
0/0/0 + 2/1 + 1/
I, 5
0/0/0 + 2/1 + 1/
0/0 + 0/2/1 + 1/
1,5
0/0/2/1 + 1/
/0+0/2/1 + 1/
//2+1 + 1 + 1 + 1/1 + 1 + 1/
1.5
0/0 + 0/P/l/
1,5
0/0 + 0/2/1 + 1/
1,5
0/0/0+1/1 + 1/
1.5
0/0/0 + 2/1 + 1/
0/0/0 + 2/1/
0/0/0/2+1/
8-10 + 8-10
9 + 8
9 + 8
9 + 8
13+12-13
8 + 7
14+12
9 + 8
9-10 + 7-9
9 + 8
6-8 + 7
9 + 8
7-9 + 7-9
9 + 8
3-5 + 3-5
9 + 8
8-13 + 8-13
9 + 8
10-13 + 9-13
9 + 8
4 + 3-4
9 + 8
7-10 + 7-10
9 + 8
6 + 5-6
9 + 8
10-13 + 9-13
JOHNSON: PERCOIDEI
481
Table 120. Continued. E,\tended.
1 \i r^Ai
SKFt I^TON
H/E U/Ah
H Fusions
Proc
spur
Pili
can.
BR
lAC
Scales
5/3/1/2
1
7
+
Ct
5/3/1/2
-(r)-
6
+
Ct
—
+
5/3/2/2
+ +
7
+
Cl
2-3/3/0/2
+
6
+
Ct
1-2; 3-4-UR
5 absent
5/3/2/2
+ +
?
7
+
Ct
5/3/2/2
?
7
+
Ct
5/3/2/2
+ +
7
7
+
Ct
5/3/2/2
+
7
+
Ct'
3/3/2/2
1-2; 3-4
+ +
?
7
+
Ct
5/3/2/2
+ +
+
6
+
Ct
5/3/2/2
+ +
+
7
+
Ct
3/3/2/2
1-2; 3-4
+ +
+
7
+
Cy
5/3/2/2
+ +
+
7
+
Ct
4/3/1/2
3-4
?
6
7
Cy
3/3/1/2
1-2; 3-4
5
-
Cy
4/3/2/2
3-1
+ +
?
6
+
Ct
5/3/2/2
?
6
+
Ct
5/3/2/2
+ -
6
+
Ct
3 or 5/3/2/2
(1-2; 3-4)
7
+
Ct
5/3/2/1-2
+
6
+
Ct
2/3/0/0/
1-2-3-4-UR
7
+
Cy
5/3/2/2
+ +
(+)
7
+
Ct
2-5/3/2/2
(1-2; 2-3; 3-4)
+ -
+
7
+
Ct
5/3/2/2
+ +
+
7
+
Ct
bones in relation to the development of the dorsal fin may prove
useful in determining the homologies of these additional ele-
ments as well as the first three predorsals.
Caudal skeleton.— The primitive percoid caudal skeleton con-
sists of one parhypural with a well-developed hypurapophysis,
five hypurals, two pairs of uroneurals. three epurals, one ural
centrum, a low neural crest on PU2 and autogenous haemal
spines on PU2 and PU3. This configuration is found in at least
some members of 54 percoid groups.
The most common reductions involve fusion of hypurals one
and two and hypurals three and four and loss of the posterior
uroneural pair. Loss of one epural occurs in only 14 groups, and
epurals are completely lacking only in some congrogadids. More
extreme reductions, including various combinations of fusions
of the hypurals with the parhypural and/or urostyle, loss or
fusion of the anterior uroneural pair and fusion of the autoge-
nous haemal spines, occur in only a few groups: Acanthoclini-
dae, some apogonids, Congrogadidae, Grammatidae, Menidae,
Mullidae, Opstognathidae, Plesiopidae, and Pseudochromidae.
The second preural centrum bears a full neural spine in only
two groups, Echneididae and Nandidae, except for occasional
anomalous specimens. This full neural spine must be second-
arily derived in the echeneidids because these fishes are un-
questionably closely related to other percoids that bear the usual
reduced neural crest on PU2 (see discussion on utility of larval
morphology). Unfortunately, evidence for the origin of this de
novo spine in echeneidids is lacking. Although it may represent
a captured first epural (there are only two in echeneidids), it is
attached and of full length at its earliest appearance in ontogeny.
Another possibility is that the second preural centrum of other
percoids has been lost in echeneidids, so that the last centrum
bearing a full neural spine actually corresponds to PU3. How-
ever, presence of the usual autogenous haemal spines on both
PU2 and PU3 in echeneidids refutes this hypothesis. The sig-
nificance of a full neural spine on PU2 in the Nandidae is un-
clear, since the affinities of this family with the Percoidei remain
problematic.
The presence of a procurrent spur and of radial cartilages
anterior to the neural and haemal spines of PU3 are probably
primitive features (Johnson, 1975, 1983). The procurrent spur
is developed to some extent in 50 percoid groups, all but ten of
which have a primitive caudal complex. Reductions among these
ten groups usually involve only simple hypural fusion. The pro-
current spur is never present in groups with fewer than 9 -F 8
principal rays. Third preural radial cartilages are found in 45
of 66 groups examined for them, about half of which have
primitive caudal complexes.
I'crtchral number. — \ c'rXf:bra.\ number ranges from 23 to about
78 in percoids. Gosline (1968, 1971) noted that the "basal
number" of vertebrae in percoids is 24-25 (10 -I- 14-15), and
this number characterizes 45 of the 91 groups treated in Table
1 20; 63 groups have 24-27 vertebrae. Twenty-two groups have
vetebral counts greater than thirty, but only five have more than
40 vertebrae. Only priacanthids and scalophagids have fewer
than 24 (10 + 13).
G///a/-c/!e5. — Primitively, percoid gill arches contain the follow-
ing elements: one basihyal, four basibranchials (the fourth car-
tilaginous), three pairs of hypobranchials, five pairs of cerato-
branchials, four pairs of epibranchials, four pairs of
482
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 120. Continued.
Anal fin
(SS)
VII-VIII
-1,8-9
I-II. 5-
8(0-1)
IX, 2
6-29
III, 16-
X, 9
III, 7-
-17(2)
-10
-8(2)
VI-VIII
, 20-22
III. 23-
X-XII,
-26 (2)
11-22
Triseg,
PTERYG,
D
— Sla\
A
Pelvic
fin
Predorsal formulae
1,5
0/0/0-1-2/1 + 1/
0/0/0-1-1/14-1/
0/0/2/ l-t-l/
1,5
0-I-0-I-0//2/1/
1,5
0/0/2-1-1/1/
1,5
0/0/0 -i-P/P/P-l-1/
1,5
/OH-O/l/l-l-l/
/O/l/H-l/
//l/l-l-l/
1,5
0/0-1-0/2/1-1-1/
0/0/2/1 + 1/
1,5
0/0/0 + 2/1 + 1/
1,5
0/0 + 0/2+1/1/
1,5
0/0/0+1/1 + 1/
0/0/0+1/1/
0/0/0+P/l + l
1,5
0/0 + 2/1/
0/0/2/1 + 1/
CA1.1DAL FIN
Pnncipal
Mullidae
10 + 14
Nematistiidae
10 + 14
Nemipteridae
10 + 14
Neoscorpis
10 + 15
Opistognathidae
10 + 15-21
11 + 16-23
12 + 18-19
13 + 18-20
Oplegnathidae
10 + 15
Ostracoberycidae
10 + 15
10 + 16
Parascorpididae
12 + 15
Pempherididae
10 + 15
Pentacerotidae
12 + 12
12+ 13
13+11
13+ 12
13 + 13
13 + 14
Percichthyidae
10-15
+
15-23
T;25-36
Gadopsis
21 + 26
Percidae
Plesiopidae
T:31-50
10 + 15
II-III, 10-20(1-2)
XI-
XII,
11-22
111
11-
18(2)
IX, 8
-10
III, 7-
8(2)
XI-
XII,
14-17
111
13-
15(2)
IV
-VII
7-12
III
17-
45(2)
IV-
-XV
,8-29
II-VI, 6-17(?-2)
VII-XI. 8-18
111,7-13(2)
X-XII, 25-28
III, 17-19(2)
V-IXX-0-lII, 7-24
1-11,4-15(1-2)
IX-XV, 7-21
+
16-25
T:26-35
I-III, 13-30
i-
^*
II
4-7
13
— +
11
0-16
0-15
^ + (-)
6-16
10 + 16
III, 8-23 (2)
6-19
12 + 25
Pomacanthidae
10 + 14
IX-XV, 15-33
0
III-IV. 14-25(2)
0
Pomatomidae
11 + 15
VIl-IX, 23-28
2-3
III, 22-28 (2)
3
Polyprion
13 + 14
XI-XII, 11-13
8-10
111,8-10(2)
5-6
Priacanthidae
10+ 13
X, 11-15
0-1
111,9-16(2)
0-1
Pseudochromidae
10-13
1-111,21-37
0-many
0-many
I, 5
variable:
0-5 predorsals
0-2 sup. spines
I, 1
//O/O/l/l/
1,5
0/1/1/1/
/l/l/l/
////I/l/
/////l/l/
1,4
0/0/0 + 2/1 + 1/
0/0/2/1 + 1/
0/0/0+1/1 + 1/
0/0/1/1 + 1/
1.5
0/0/2/1 + 1/
0//2/1 + 1/
1,5
0/0/0+1/1 + 1/
I, 5
0/0/0+2/1 + 1/
1,5
0 + 2/1/1/1/
2/1/1/1/
.3-5
0/0/0 + 2/1 + 1/
0/0/0+1/1 + 1/
0/0/2/1 + 1/
0/0/1/1 + 1/
8 + 7
8-10+8-10
9 + 8
9-10 + 8-9
9 + 8
8-11+8-11
9 + 8
9 + 8
6-7 + 6-7
3-8 + 3-7
9 + 8
9 + 8
9 + 8
10-12+10-11
9+8
11+9
9 + 8
3-7 + 3-7
9 + 8
3-7 + 3-6
9 + 8
5-
16 + 5-
9 + 8
5 + 5
-14
8
-9 + 7-
-8
10-
15 + 8-
-13
9-
10 + 8-
-9
3-10 + 3-9
9 + 8
4 + 3-4
9 + 8
9-10 + 8-9
9 + 8
8-9 + 7-8
8 + 8
4-6 + 4-6
7-9 + 7-8
5-7 + 5-7
JOHNSON: PERCOIDEI
483
Table 120. Continued. Extended.
( M DM
SKI 1 t ION
H h I 1 Ah
H Fusions
Proc
cart
BR
lAC
Scales
3/2/1-2/2
1-2; 3-4-UR
?
4
-
Ct
3/3/2/2
1-2; 3-4
7
+
Cy
5/3/2/2
+
6
r
or
Ct
5/3/2/2
+ +
6
+
Ct
2-3/3/0/1
+
6
+
Cy
pH-1-2; 3-4-UR;
(5 absent)
5/3/2/2
+ +
?
7
+
Ct
5/3/2/2
+ +
+
7
+
Ct'
5/3/2/2
+ +
7
+
Ct
3 or 5/3/1-2/0-2
-( + ) +
7
+
Ct
(1-2; 3-4)
+
or
Cy
5/3/1/2
+ -( + )
7
+
Ct
_
+
pharyngobranchials, and an interarcual cartilage between the
uncinate process of epibranchial 1 and pharyngobranchial 2.
The first pharyngobranchial is rod-like and serves to suspend
the dorsal gill arches from the neurocranium. The fourth phar-
yngobranchial is reduced and cartilaginous, but consistently bears
a well-developed dermal tooth plate, as do the second and third
pharyngobranchials and the fifth ceratobranchials. Small tooth-
plates on the second and third epibranchials are variously pres-
ent or absent.
Reductive departures from the primitive branchial complex
are few and involve only the basihyal, first pharyngobranchial
or interarcual cartilage. The basihyal is reduced or absent in
ephippidids. Pseudochromids lack a first pharyngobranchial
(Springer et al., 1977). Of 88 percoid groups examined for it,
only 1 3 lack a well-developed interarcual cartilage and at least
three of these (Cirrhitidae, Emmelichthyidaeand Nemipteridae)
may have a vestigial element. The remaining eleven groups
completely lack the interarcual cartilage, but most have an un-
cinate process with the cartilaginous tip separated by a decided
gap from the second pharyngobranchial and frequently pointing
away from it. This condition differs from the primitive state (as
represented in most beryciforms) wherein the uncinate process
of the first epibranchial directly contacts that of the second
pharyngobranchial, and suggests that these percoids have sec-
ondarily lost the interarcual cartilage. A condition resembling
that of the beryciforms was observed among percoids only in
some anthiin serranids, where it must be secondary. In eche-
neidids the uncinate process of the first epibranchial also artic-
ulates directly with that of the second, but there is a concom-
itant extreme reduction of the main arm of the first epibranchial
not seen in beryciforms. Again this condition must be derived
if the relationships of the echeneidids are as postulated here (see
discussion on utility of larval morphology).
7/2-3/1-2/2
+
+ (- -)
5-7
+
Ct
■>
+ (-)
or
Cy
5/2/1/2
7
+
Cy
5/3/1/2
5-8
+
Ct
3/3/0-2/1
l-2-(pH); 3-4-UR
5/3/2/2
5/3/2/2
5/3/2/2
5/3/2/2
3/2-3/0/0-1/
(pH)- 1-2; 3-4-UR
+ (-)
—
6
+
Ct
+
or
6
+
Ct'
-1-
+ +
7
+
Cy
-1- +
7
+
Ct
7
6
-
Ct'
_ _
6
_
Ct
Scales.— The unpublished work of McCully (1961) on compar-
ative anatomy of serranid scales provides an excellent illustra-
tion of the wealth of information available in the scales of per-
coid fishes that has largely been ignored in systematic studies.
More recent work on ctenoid scales of other groups (DeLamater
and Courtenay, 1973a, b, 1974; Hughes, 1981) using scanning
electron microscopy also demonstrates the systematic value of
ctenoid scales. Details of the scale morphology of most percoids
are unknown. On a gross level, three basic scale types (Ct, Ct'
and Cy in Table 120) are found among percoids. Although be-
ryciforms and some myctophids are said to have ctenoid scales,
these scales (Ct') differ from the type possessed by most percoids
and other perciforms (Ct). In beryciforms and myctophids the
"ctenii" are continuous spinous projections from the lateral sur-
face and posterior margin of the scale. A few percoids (Bramidae,
Epigonidae, Howella. Pomacanthidae, Priacanthidae, Ostraco-
berycidae and Scatophagidae) possess similar scales that may
represent retention of the plesiomorphic beryciform condition,
or may have been secondarily derived. In the "true" ctenoid
scale that characterizes most percoids (59 groups), the ctenii are
separate bony plates, or scalelets (McCully, 1961, 1970), that
are continually added in the posterior field as the scale grows.
In most groups the posterior field becomes filled with remnants
of old ctenii, the tips of which are amputated (or, more likely,
resorbed), as each new row of ctenii is added. In a few groups,
however (e.g., anthiine serranids and callanthiids), only a pri-
mary and secondary row of marginal ctenii are evident. This
second variation of "true" ctenoid scale also characterizes
484
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 120. Continued.
Triseg.
rALIDAL I-TN
Vertebrae
Dorsal tin
PTERYG.
D
— Slav
A
Pelvic
fin
Predorsal formulae
Pnncipal
Anal fin (SS)
Procurrent
Rachycentridae
11 +
14
VIl-IX, 26-34
1-11,22-28(1)
0
0
I, 5
/l + l/l/l/
9 + 8
15-16+12-14
Scatophagidae
10 +
13
Xl-Xll, 16-18
IV, 14-16(2)
0
0 +
I, 5
0/0 + 2/1/1/
0 + 0/2/1/1/
8 + 8
4-6 + 4-5
Sciaenidae
10-
+
5
VII-XV-1, 17-46
1-11,5-23(1-2)
l^
1,5
0/0/0 + 2/1 + 1/
//2+l + l + l,etc./
9 + 8
7-10 + 6-9
12-
S
T:24-
-29
Scombropidae
10 +
16
VIlI-lX-1, 12-13
h
1,5
0/0/0 + 2/1 + 1/
9 + 8
II, 11-12(1)
11 + 10
Scorpididae
10 +
10 +
15
16
IX-X, 22-28
III. 17-28(2)
m^
1, 5
0/0/0 + 2/1 + 1/
9 + 8
11-13+10-12
Serranidae
10 +
11 +
14
13
VI-XIl, 9-24
or
0-24
0-19 "^
I, 5
0/0/0 + 2/1 + 1/
0/0 + 0/2/1 + 1/
8-9 + 7-8
3-12 + 3-10
10 +
15
II-IV, 20-29
0/0/2/1 + 1/
10 +
16
II-III, 6-22 (2)
0 + 0/2/1/1/
11 +
15
or
0/0/1/1 + 1/
10 +
18
13-17(1)
0/0/P/l + l/
0//P/1 + 1/
//l/l + l/
Sillaginidae
14-20
+
X-XIII-I, 16-27
11, 14-26
^
I, 5
0/0/0/1 + 1/
0/0/0/1/
9 + 8
17-19+14-19
19-27
0//0/1/1 + 1/
T:33-
-44
Siniperca
12 +
13 +
16
15
XI-XV, 10-17
111,7-13(2)
7-10
4-6
1,5
0/0/0 + 2/1 + 1/
0/0/0/2/1 + 1/
9 + 8
6-12 + 6-12
13 +
18
Sparidae
10 +
14
X-XIII, 10-15
111,7-14(2)
1-4
1,5
0/0+0/2+1/1/
9 + 8
7-11+7-11
Stereolepis
12 +
14
XI-XII,9-10
III, 7-9 (2)
W
1.5
0/0+0/2/1 + 1/
0/0/0 + 2/1 + 1/
9 + 8
10-11+8-9
Symphysanodon
10 +
15
IX, 10
w
1,5
0/0/0 + 2+1/1/
9 + 8
III, 7-8 (2)
12-14+12-14
Teraponidae
10 +
10 +
15
16
XI-XIV, 8-14
111,7-12(2)
0
0 +
1,5
0/0+0/2/1 + 1/
0/0+0/1/1 + 1/
9 + 8
9-10 + 5-8
10 +
17
0+0/0 + 2/1/1/
11 +
14
11 +
16
Toxotidae
10 +
14
IV-Vl, 11-14
III, 15-18(2)
0(+)
0 -
I, 5
0/0/0//P/P/1/
0/0/0//P/1/
9 + 8
4-5 + 4-5
0//0//P/P/1/
Aphredoderus. gobies and some flatfishes, and the mechanism
of growth of the posterior field is not understood. As shown by
the authors mentioned above, there is extensive diversity in
configuration and processes of formation of marginal and sub-
marginal ctenii, and this diversity undoubtedly holds useful
phylogenetic information.
The third major scale type found among percoids is the cy-
cloid scale (Cy in Table 1 20), characteristic of most groups below
the Percomorpha. Although the cycloid scales of some percoids
may represent a plesiomorphic state, they are clearly secondary
in a number of families where they occur only in some members
(acanthoclinids, acropomatids, apogomds, ephippidids, pem-
pheridids, percichthyids, sciaenids and serranids). Cycloid scales
also characterize all members of two groups of percoid families.
each of which probably comprises a monophyletic lineage. The
cirrhitoid fishes (Aplodactylidae, Cheilodactylidae, Chironem-
idae, Cirrhilidae, and Latrididae) have large cycloid scales of
similar morphology, and the carangoid fishes (Carangidae, Cor-
yphaenidae, Nematistiidae, Rachycentridae and Echeneididae)
have very small adherent cycloid scales. Cycloid scales char-
acterize five other families of moderate size, Ambassidae, Ce-
polidae. Congrogadidae, Leiognathidae and Opistognathidae.
Otherwise, cycloid scales are restricted to a few monotypic fam-
ilies and tncertae sedis genera (Bathyclupeidae, Caristiidae, Di-
nolestidae, Drepamdae, Enoplosidae, Lactanidae, Menidae, Po-
matomidae, Scombropidae and Siniperca). The widespread
occurrence of true ctenoid scales in the Percoidei, including most
of the less specialized forms, and the distribution of cycloid
JOHNSON: PERCOIDEI
485
Table 120. Continued. Extended.
t AliDAl
Proc
SKELETt)N
H/E/U/Ah
H Fusions
spur
cart.
BR
lAC
Scales
5/3/2/2
h
7
+
Cy
5/3/1/2
6
-
CV
5/3/2/2
+ +(--)
7
+
Ct
—
?
or
Cy
5/3/2/2
+ +
+
7
+
Cy
3 or 5/3/2/2
(1-2:3-4)
+ +
(+)
7
+
Ct
3 or 5/3/1/2
(1-2; 3-4)
-
6-7
+
(-)
Ct
or
Cy
5/2-3/1-2/2
+ +
+
6
+
Ct
4-5/3/1/2
r(-) +
7
+
cv
(3-4)
-
3 or 5/3/2/2
(1-2: 3-4)
+
6
+
Ct
5/3/2/2
+ +
7
+
Ct
5/3/2/2
- +
+
7
+
Ct
5/3/2/2
+ +
+
6
+
Ct
5/3/0-1/2
+ -
Ct
fully coalesced by hatching. Most members of the three primary
freshwater families. Centrarchidae, Percichthyidae and Percidae
have demersal eggs as do some members of the Ambassidae
and Teraponidae, however only six families of exclusively ma-
rine percoids are known to possess non-buoyant eggs. The Acan-
thoclinidae. Congrogadidae, Plesiopidae and Pseudochromidae
have specialized demersal eggs with adhesive threads that bind
them together in attached, sheet-like (Plesiopidae) or free, spher-
ical (Acanthoclinidae and Pseudochromidae) masses that are
guarded by the male. These eggs also have numerous small oil
globules that gradually coalesce with a single, much larger glob-
ule. The possibility that these four families are closely related
has remained unresolved (Bohlke, 1960a; Springer et al., 1977),
but the similar egg morphology and parental care shared by
them may represent synapomorphies not heretofore considered.
The other two marine families with adhesive demersal eggs,
Apogonidae and Opistognathidae are oral brooders, and oral
brooding has also been reported for the plesiopid Assessor (Allen
and Kuiter, 1976).
Larvae
Diversity of general body form and morphological special-
ization among the larvae of percoid fishes is extensive, and. as
with the adults, no single feature shared by larval percoids char-
acterizes the suborder. Representative postflexion larvae of 62
percoid groups are illustrated in Figs. 254-262. Larval serranids
and carangids were excluded from these figures because they are
illustrated elsewhere in this volume. I was unable to obtain
specimens or illustrations of larvae of the remaining 30 groups
and most are probably unknown, or at least undescribed. Of
these, 19 are monotypic.
Larval body form ranges from elongate to deep-bodied, by
the criteria of Leis and Rennis (1983), and frequently, but not
always, reflects adult body form. Thus, some of the most deep-
bodied percoid larvae are found among the Chaetodontidae,
Pomacanthidae, Menidae, Bramidae, and Caristiidae, whereas
the elongate Congrogadidae and Cepolidae have elongate larvae.
On the other hand, the moderately elongate larvae of groups
like the Girellidae or the Cirrhitidae are not particularly reflec-
tive of the adult body form, nor are the deeper-bodied larvae
of the Emmelichthyidae.
In Table 121, selected aspects of known larvae of percoid
families and mcertae sedis genera are given. This table should
prove a useful guide to identification of postflexion larval per-
coids at the family level, particularly when used in conjunction
with the meristic data in Table 1 20 and the illustrations in Figs.
254-262. Features included in Table 121 are discussed below.
scales just described, suggests that cycloid scales in most per-
coids have been secondanly acquired.
Development
Eggs
Most percoids have buoyant, spherical eggs about 1mm in
diameter, with a single oil globule. The total size range is about
.5 to 4.6 mm, but eggs larger than 2 mm are found only in a
few freshwater-associated groups, Centrarchidae, Moronidae.
Percichthyidae, Percidae, Siniperca and Teraponidae, and in
the marine Echeneididae (Table 121). Multiple oil globules oc-
cur in some centrarchids, percichthyids and sciaenids, and in
Hapalogenys. moronids and Polyprion, but they are generally
Fin development. — Formation of median fin rays occurs at very
small sizes in most percoids. Flexion may begin as early as 2.5-
3 mm and is complete in most groups by 4-5 mm, at which
time the full complement of principal caudal rays is present.
Dorsal and anal fin rays begin to form during or shortly after
flexion and are usually complete, including spinous rays, by 5-
8 mm. Size at flexion and completion of full median fin ray
complements is relatively consistent within families, the range
usually not varying more than 2 mm. Groups characterized by
notably later flexion (6-18 mm) include the Caristiidae, Cen-
tracanthidae, Centrarchidae, Cheilodactylidae, Girellidae, La-
tcolabrax, Moronc. Percichthyidae, Percidae, Polyprion. Scor-
pididae, Sillaginidae, and Siniperca. These groups also exhibit
somewhat delayed dorsal and anal fin ray completion (7-18
mm). Among marine percoids, the most extreme delay in com-
486
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 121. Selected Early Life History Features of Percoidel Parentheses enclose features known to characterize only some members of
a group. Head spination abbreviations— Supraoccipital: SI —small peak-like crest; S2 — SI with serrations; S3 — large vaulted spine-like crest with
serrations; S4 — low serrated median ridge; S5— entire surface rugose. Frontal: Fl —entire surface rugose; F2— one or more parallel or converging
serrated ridges; F3 — serrated supraorbital ridges; F4 — single spine on supraorbital ridge; F5 — large posteriorly projecting serrated spine. Preopercle:
PI— posterior margin with moderate to large simple spines; P2 — PI plus lateral ridge with one or more small simple spines; P3 — P2 with spine
at angle notably elongate; P4 — P3 with marginal spines serrate; P5 — posterior margin and sometimes lateral ridge with very small spines or
serrations. Other bones with simple spines, serrations or serrated ridges: Op— opercle; Sb — subopercle; lo— interopercle; Ta— tabular; Pt —
posttemporal; Scl — supracleithrum; CI— cleithrum; La — lacrimal; Co— circumorbitals; Na— nasal; Mx — maxillary shaft; D— dentary; Br— bran-
chiostegals; Pe — pterotic; Pa — parietal; Sp— sphenotic. Sequence of completion of fin rays: A. D,-A-D|-P,-P|; B. D,-P,-D,-A-P,; C. P,-P,-D,-A-
D,; D. P|-D,-A-D|-P,; E. A-D,-P,-P|-Di; F. P,-D,-A-D|-P|. Egg type: P— pelagic, buoyant; D— demersal; A — adhesive; M — egg mass; O— oral
brooder.
Size (mir
)
Sequence
Text
D & A rays
First
of fin
Head
Other
Taxon
figures
Egg type
Egg
Halch
Flex
complete
scales
completion spination
specializations
Acanthoclinidae
255D
D, A, M
-1.4
-4.7
5-6
7
7
A
P5
None
Acropomatidae
254A-D
P
?
9
-4
-5
12-15
A
(SI),(S4),(F2),
F3, (P4), (P5).
Op, Sb, lo, Pt
Scl, (Pel),
(Co), (D), (Pe)
(D and P,
spines ser-
rate)
Ambassidae
255A
(D, A)(P)
.7-. 8
1.8
-3.5
5.5-6
9-10
A
P5
None
Apogonidae
257E>-G
D, A, M, O
<1
2.5-3
3-4
4-6
12 or >
A(B)
(S1),(S5),(F1),
(P2), (P3),
(P5), (Op),
(Sb), (Pt)
(. . .?)
(Fl), PI,Op,
(Elongate D
spines and
P, rays)
Bramidae
261E
P
9
-3
4-7
6-10
7-10
C(D)
Spinous scales
Sb, lo
(large P, and
Callanthiidae
255E-F
P
?
7
5
7
7-14
A
P2, Op, Sb, lo,
Pt
(S1),(F3),(F4),
None
Carangidae
P
.7-1.3
1-3.5
-3-5
-6-10
-7-14
A(D)
(Elongate D
P3, (P4), Pt,
spines and
Scl, (Pe)
P2 rays)
Caristiidae
261D
P
1.1-1.3
2.3-2.9
-7
-8
7
A
P5
None
Centracanthidae
258J
P
1.1-1.3
2.3-2.9
6-7
8-9
7
A
P5
None
Centropomidae
260G
P
.7
1.4-1.5
3.6-3.8
-7
-14
A
P5
None
Centrarchidae
260A
D, A
.8-2.8
2.2-5.5
6-9
-7-13
-14-18
A
None
None
Cepolidae
262G
P
.7
<3
8-9
7-9
7
A
S3, F1,F3, P4,
Scl, D
None
Chaetodontidae
262A-C
P
.7-.9
1.5-2.0
4-5
5-8
7-11
A(B)
All exposed
head bones
thick and ru-
gose.
Pt and Scl ex-
panded poste-
riorly.
P expanded to
cover cheek
and with
broad flat
spine poste-
riorly.
(P, spine long
and serrate)
(Ant. D spines
long and ru-
gose)
Cheilodactylidae
258E
P
.9-1
2.9-3.3
7-8
10-12
-10
A
None
Postlarvae
deep, com-
pressed, sil-
very to 70-
90 mm
Cirrhitidae
258F
P
?
7
-4
-8
10 or <
A
P5
Chin barbel
Congrogadidac
255G
D, A, M
7
7
7
7
7
7
None
None
Coryphaenidae
26IA
P
-1.6
-4
6.5-7.5
D 1 3-24
A 8-11
-25-30
E
F4, P2, Pt
Minute epithe-
lial "prick-
les" by
-6 mm;
"swollen"
pterotics
Echeneididae
261C
P
1.4-2.6
4.7-7.5
5-9
D 12-30
A 6-12
- 1 5-30
E
None
Large hook-
like teeth on
dentary
Emmelichthyidae
2591
P
?
7
?
7
7
A
Pl,Op, lo, Pt,
Scl
None
JOHNSON: PERCOIDEI
Table 121. Continued.
487
.Size (mm)
Spfiiipnr*'
Text
D & A rays
First
of fin
Head
Other
Taxon
figures
Egg lype
Egg
Hatch
Flex
complete
scales
completion spination
specializations
Ephippididae
Chaetodipterus
256G
P
~l
-2.5
-4
-5
-8-9
A
S1,F3. P2. Op,
lo, Ta, Pt
Spinous scales
to - 1 5 mm
Epigonidae
Epigonus
257B
P
?
7
7
7
7
7
None
None
Sphyraenops
257A
P
7
7
7
7
-12
7
S1,S5, F1,F3,
None
P3, Op, Pt, Pe
Gerreidae
259A
P
.6-. 7 5
-1.4
3.5-4.4
-6
>15
A
P5, (Scl)
None
Girellidae
258C
P
~1
-2.3
-6
11-13
-15-16
A
P5, Scl
None
Haemulidae
259B-D
P
.8-1.0
1.7-2.8
3.9-5.4
6-8
(earlier
in
P. nigra)
-13
(much
earlier m
Conodon)
A
(F3), (P1),(P5),
(Op), Sb, lo,
(Pt), Scl, Pe
(also F2, Pel,
La, Co, Na, D
in Conodon)
None (spinous
scales in
Conodon)
Hapalogenys
254H
P
1.2
-3
-4.5
-5-6
>10
F
S3. S5,FI,F3,
P2, Op, Sb,
lo, Pt, Scl, La
Pe
P, preco-
cious, large
Howella
257C
P
?
7
-3.5
-4.5
7
A
P5, Op, lo, Pt,
CI
P5. Op, Sb, lo.
None
Kyphosidae
259J
P
~1
2.4-2.9
3.8-5.5
6-7
-7
A
Spinous scales
Scl, Pel
Lateolabrax
260E
P
1.3-1.4
4.4-4.6
-9
-15
>I5
A
P5
None
Leiognathidae
256C
P
.6
1.4
-4
-5
7
7
S3, F3, P4, Pt.
Scl
Ant. D spines
serrate
Leptobramidae
258D
7
?
7
9
>8
>9
7
None
None
Lethrinidae
262F
P
.7-.8
1.3-1.7
4.4-5.2
5.5-7.0
8
A
S3, F3, P4, Op,
Sb, lo, Ta, Pt,
Scl. Pel, U,
Co, Mx, D,
Pe
Spmous scales
Lobotidae
254G
P
~1
7
<6
<6
-8
7
S3, S5, FI,F4,
P2, Op, Sb,
lo, Pt, Scl
P, large (pre-
cocious?)
Lutjanidae
256A-B
P
~.5-.8
1.7-2.2 4.2-5.3
5-6
-12-14
B
F3, (P2). (P3),
(2nd dorsal
(including
(P4), Op, lo.
spine and P,
Caesionidae)
Pt, Scl. Pel
spine and
soft rays
elongate)
(anterior D,
A, and P,
spines ser-
rate)
Malacanthidae
256E-F
P
1.2
2.2-2.6
4-6
5-8
3-4
A
F2, F3, P4, Op,
Sb, lo, Ta, Pt,
Scl, La, Na,
D, Pe, Pa
Spinous scales
to -30 mm
or >; fused
nasals
Mcnidae
256D
7
?
7
<4.5
<4.5
7
9
None
None
Microcanthidae
259G
P
7
7
-4
5-6
-15
A
P2, Op, Sb. lo,
Pt, Scl
None
Monodactylidae
255H
P
.6-.7
1.8
3.5-4.0
5-6
9-10
F
F3, P2. Op, lo,
Pt, Scl
P, large, pre-
cocious
Moronidae
260F
(P) (D, A)
.7-4.6
1.7-3.7
7-9
10-13
16-25
A
P5
None
Mullidae
259E
P
.6-.9
1.6-3.4
3.5-4.5
-7
-12-13
A
None
Silvery, pelag-
ic postlarvae
to -40-
60 mm
Ncmipteridae
258H
P
.7-.8
1.5-1.6
-4
6-8
-11
A
None
None
Opistognathidae
255B
D, A, M. O
7
9
-5.5
-7
9
A
P5, lo
None
Oplegnathidae
255J
P
.9
2.3
-5
-7
-12
A
P5, Op, lo, Scl
None
Pemphendidae
2551
P
7
7
3.6-4.3
-6
7
F
P5, lo, Scl
P, precocious
Pentacerotidae
262J
P
?
7
7
7
-12
7
S3, S5, FI,F3,
F4, P4, Op,
Pt, CI, La, Pe
P, spines ser-
rate; spinous
scales
Percichthyidae
260D
(P)(D, A)
1.2-4.2
3.1-9.0
7-9
9-13
10-20
D
None
None
488
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 121. Continued.
Size (mm)
Text
hgures
Egg type
Egg
D & A rays
Hatch Flex complete
First
stales
Sequence
of fin
completion
Head
spinalion
Other
specializations
Percidae
Plesiopidae
Polypnon
260C
254E
(P)(D, A)(M)
D, A, M, (O)
P
.7-2.8
-.9 X .6
1.6
Pomacanthidae
256H
P
.7-.9
Pomatomidae
Priacanthidae
258G
262H-I
P
P
.8-1.2
9
Pseudochromidae 255C
Rachycentridae 26 IB
Scatophagidae
262D
Sciaenidae
257H
P
Scombropidae
262E
P
Scorpididae
258A-B
P
Serranidae
Serraninae
-
P
Anthiinae
P
Epinephelinae
Epinephelini
Grammistini
4.7-8.7 7-15 9-18
2.8-2.9 ? ?
3.7 -7 -9
1.5-1.8 3.4-4.3 4-5
2-2.5 5-6 ~7
? 4-5 ~7
.7-1.0
~4
-4-5
1.4-2.4 -4-5
<2.9 3.3-4.6 -6
13-24
2.5-2.8
-12
-6
D. A, M
?
3-4
4.4-5.8 -8
12-13
P
1.2-1.4
9
-7 D 16-18
A 9-10
30-35
-4
7-1.3
1.5-2.5 3.0-4.6
-5-9
14-20
?
? ?
>6
9
?
? -6
9-10
9-11
8-1.0
-2.2 4.3-5
~6
-11-12
6-. 8
1.2-1.4 3.5-5
4.6-5.5
~6->l
>15
A(D)
(P5)
None
9
9
9
A
S2, F3, P2, Op,
Sb, lo, Pt, Scl
None
A
F3, P2, Sb, lo.
Spinous scales
Ta. Pt, Scl,
to 17-
La, Co, Na, D
19 mm
A
P5
None
A
S3, F2, F3, P4,
D„A, P,
Op, Sb, lo.
spines and
Ta, Pt, Scl,
soft rays ser-
La, Co, Na,
rate; spinous
D, Br
scales to
— 20 mm
A
P5
None
E
F4. P2, Pt
Minute epithe-
lial pnckles
by ~6 mm
"swollen"
ptenotics
D
Most exposed
head bones
thick and ru-
gose; P and
Pt expanded
posteriorly; Pt
with posterior
spatulate
"spine"; Pe
swollen and
with separate
rugose
"shield"
Spinous scales
A(B)
(S4), (F3), (P2),
P5, Pt, Scl, lo
None
A
S2, F3, P4, Op,
Sb, lo. Pt
None
A
P2, Op, Sb, lo,
Scl
None
A(B)
P2, Op, Sb. lo,
Pt, Scl
None
B(A)
(SI),(S2),(F1),
(D,. A and P,
(F2), (F3),
spines ser-
(F4). (P3),
rate) (ant D
(P4), Op,
spines and
Sb, lo, (Ta).
P., rays elon-
Pt, Scl, (La).
gate)
(Co), (D),
(Pe), (Pa)
B
(Fl), F3, P4,
D,, A and P^
Op, Sb, lo, Pt,
spines ser-
Scl
rate; second
D| spine
and P,
spines elon-
gate
P„D„D„
P2, Op, Sb,
Ant D spines
A,P2
lo
flexible,
elongate,
pigmented;
P, large,
precocious
JOHNSON: PERCOIDEI
489
Table 121. Continued.
Text
5gures
Egg type
Size (mm
Head
1 spmation
Taxon
Egg
Hatch
Flex
D & A rays
complete
First
scales
of fin
completio
Other
specializations
Liopropomini
P
~1
7
9
~6
7
D„D„A.
P„P,
(F4). P2, Op,
do)
Ant D spines
flexible,
elongate, or-
namented
Sillaginidae
259F
P
.6-.7
1.3
-6
-9
9
A
P5, Pt
None
Siniperca
260B
D, A
-2
~5
-10
-11
9
A
P5
None
Sparidae
2581
P
.8-1.2
2.0-2.7
4-7
6-11
8-20
A
P5. lo, Pt. Scl
(also S2, F3,
None (spinous
scales in Pa-
Stereolepis
254F
P
?
7
7
<7
>10
7
P2, in Pagrus
F3. P2, Sb. lo.
Pt. Scl, Pe
grus)
None
Symphysanodon
254A
P
?
7
3.5-4.0
-4.5
-13-14
A
F2, F5, P4, Pt.
Scl, Ta, La,
Co, D, Pe
None
Teraponidae
259H
(P)(D)
.7-2.8
1,7-3.7
-4-8
-7-11
14-18
A
P5, Op, Sb, lo,
Scl, CI, Pel
None
pletion of dorsal fin rays (12-30 mm) occurs in the elongate
larvae of Coryphaena. Rachycentron and the Echeneididae.
The most commonly observed sequence of fin completion
(pattern A in Table 1 2 1 ) is that described for Moronehy Fritzsche
and Johnson (1980) and for Anisolremus by PotthofTet al. ( 1 984).
Soft rays of the dorsal and anal fins begin to form during or just
prior to flexion. Fin rays appear first near the future middle of
these fins and are added in an anterior and posterior direction.
Full complements of dorsal and anal soft rays are usually achieved
at about the same time as the full principal caudal fin ray com-
plement. The spinous dorsal fin is completed next (usually from
posterior lo anterior) followed by the pelvic and pectoral fins.
Precocious development of the anterior portion of the spinous
dorsal and the pelvic fins, pattern B, is usually associated with
ornamentation and/or elongation of the spines. It characterizes
all larvae of lutjanids and epinepheline serranids, and a few
apogonids, chaetodontids and sciaenids. In liopropomine ser-
ranids, the anterior portion of the spinous dorsal is precocious,
but the pelvic fins develop last. Precocious development of pec-
toral and pelvic fins, pattern C, is unique to some members of
the Bramidae. Pattern D, precocious pectorals only, is found in
scatophagids, some bramids, and interestingly, is also shared
by the freshwater Percichthyidae and some Percidae. The pec-
toral fin and anterior portion of the spinous dorsal are precocious
in the serranid tribe Grammistini. In pattern E, the full anal fin
ray complement tends to be complete prior to that of the dorsal,
and the spinous dorsal is the last fin to be completed. This
pattern is unique to the echeneoid fishes (Coryphaenidae, Ra-
chycentridae and Echeneididae). Pattern F, in which only the
pelvics are precocious, is found in Hapalogenys. Monodactyl-
idae and Pempherididae.
Scales. — MoiX percoids begin to develop scales well after com-
pletion of fins near the end of the larval period, frequently after
settling. In several families (e.g., Chaetodontidae, Cheilodac-
tylidae, Cirrhitidae, and Scorpididae) unspecialized scales first
appear at or slightly before completion of the median fins and
are thus present during the late larval stages. Larvae of a few
groups are characterized by early development of specialized
spinous scales that eventually transform into the typical adult
ctenoid scale. In the ephippidid Chaetodiplerus. the haemulid
Conodon. malacanthids, pomacanthids and scatophagids these
consist of small, roughly circular, non-imbricate bony plates
from the center of which one to several spines project outward
at right angles. Larvae of the Bramidae, Kyphosidae, Pentacer-
otidae, Priacanthidae, some anthiin serranids, the sparid Pagrus
and the sparoid family Lethrinidae possess spinous scales in
which one or more spines project outward at less than right
angles from the posterior field or margin of imbricate plates that
more closely resemble scales of the adults. Among non-percoid
fishes, spinous larval scales occur in trachichthyids, chiasmo-
dontids, acanthurids, Xiphias, Anttgoma and some pleuronec-
tiforms, telraodontiforms, scorpaeniforms and gasterostei-
forms. The function of specialized larval scales is unknown, but
it seems likely that they provide some defense against small
biting predators, parasites and/or nematocysts.
Head spination.— The simple to elaborate spinous ornamenta-
tion of various bones of the head in larvae of many percoid
fishes is an area ripe for future detailed investigations. Nowhere
is the potential utility of larval morphology in phylogenetic
studies more evident, for it is in this feature that larval percoids
frequently exhibit far more complexity and diversity than adults.
Although more work is needed to determine if patterns of head
spination will prove useful in studies of interfamilial relation-
ships, there can be no doubt that the diversity of these patterns
within some well-defined families or subfamilies (e.g., anthiin
serranids, chaetodontids, priacanthids, malacanthids, poma-
canthids. haemulids. etc.) offer critical information for intra-
familial phylogenetic analyses.
Extensive head spination appears to have arisen indepen-
dently numerous times within the Percoidei. Nevertheless, an
ordered progression of increasing complexity is evident in the
sequence in which ornamentation is added to various bones.
Most families are characterized by a single level of complexity,
but some are more diverse. In the larvae of several unrelated
families (e.g., Cheilodactylidae, Echeneididae, Menidae, Mul-
lidae, Percichthyidae) head spines are completely lacking. A
490
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
somewhat larger, equally heterogeneous assemblage of percoid
groups (including the Ambassidae, Centracanthidae, Centro-
pomidae. Cirrhitidae, Moronidae, Percidae, Pomatomidae and
Pseudochromidae) has minimal head spination, consisting of
only a few small spines along the posterior, and usually lateral,
margins of the preopercle. In most instances, these spines are
so small and isolated that it is difficult to imagine that they serve
any useful function.
The most common pattern of head spination among larval
percoids is one in which, in addition to small to moderate pre-
opercular spines, small spines may also occur on other bones
of the opercular series (interopercle, subopercle and opercle) and
on various bones of the pectoral series (cleithrum, postcleith-
rum, supracleithrum, posttemporal and tabulars). This pattern
occurs in many of the more generalized families that have usu-
ally been considered "basal" percoids, including the Acropo-
matidae, Gerreidae, Girellidae, Haemulidae, Kyphosidae,
Sciaenidae, Scorpididae, Sparidae and Teraponidae, and it must
be primitive for at least some large subgroup of percoid families.
Two additional levels of complexity in this artificial hierarchy
involve modifications of cranial bones (frontal and supraoccip-
ital) in addition to opercular and pectoral series spination. Mod-
ifications of the frontal bones occur only in those larvae with
opercular and pectoral series spination and encompass several
types of ornamentation. Frontal surface rugosity is found in a
few apogonids, bramids and serranids as well as in Acantho-
cepola, Lobotes, Hapalogenys, Pseudopenaceros and Sphyrae-
nops. Johnson and Keener (1984) noted this condition in larval
Alphestes. but it was not previously considered in descriptions
of percoid larvae. With closer examination, cranial rugosity will
undoubtedly be detected in larvae of other percoid and non-
percoid groups. It probably offers an efficient way to strengthen
the neurocranium during early development. Frontal spines or
serrations are most frequently borne along the supraorbital ridge.
Coryphaena, Rachycentron, Lobotes. and some carangids have
one large, broad-based supraorbital spine, but the more com-
mon condition is a series of supraorbital spines or serrations.
These are found in lutjanids, malacanthids, monodactylids, po-
macanthids, Stereolepis. some acropomatids, carangids, hae-
mulids, sciaenids, and serranids as well as in most groups with
supraoccipital modifications. More elaborate ornamentation,
consisting of a series of parallel serrated ridges on the dorsal
surface of the frontals, characterizes larval malacanthids, pria-
canthids, Synagrops and some anthiin serranids.
The most extreme example of frontal spination is seen in
Symphysanodon (Fig. 254A). A longitudinal serrated crest above
the supraorbital ridge on each frontal bone continues posteriorly
as a long, spike-like serrated spine extending to about the middle
of the spinous dorsal fin. The only other example of large paired
cranial spines among larval perciforms is found in istiophorids,
where the spines originate from the pterotics. This "homed"
effect occurs elsewhere in larvae of many scorpaeniform groups
(e.g., Scorpaenidae and Triglidae) and in the beryciforms, Di-
retmus and Anoplogaster, but in these groups the large paired
spines are parietal in origin. With the exception of occasional
minute spines or small ridges, larvae of perciform fishes never
develop parietal ornamentation, and it is tempting to speculate
that the presence of variously developed parietal spines among
larvae of many scorpaeniform groups offers support for the often
questioned monophyly of the Scorpaeniformes. In any case, this
uncommon feature should be examined in future considerations
of higher relationships among acanthopterygian fishes. The
monophyly of the Beryciformes has recently been questioned
(Zehren, 1979), and it is interesting to note that although Di-
relmus. Anoplogaster and at least some trachicthyoids share
larval parietal spines with scorpaeniforms, holocentrids lack
them, instead possessing frontal, supraoccipital and preoper-
cular spination similar to that seen in more elaborately orna-
mented larval percoids.
Modifications of the supraoccipital, representing the last cat-
egory of complexity in head spination, occur in those larvae
which also have opercular series, pectoral series and frontal
ornamentation. Simple forms of supraoccipital ornamentation
include a small peak-like median crest (Chaetodipterus, Pagrus,
Polyprion, Sphyraenops, and some acropomatids, apogonids,
carangids and anthiin serranids) or a serrated, ridge-like crest
(Synagrops. some sciaenids and anthiin serranids). The more
extreme form is a large, vaulted, variously serrate spine-like
crest that projects beyond the posterior margin of the cranium
and is well-developed in preflexion larvae soon after hatching.
This type of crest characterizes larval cepolids, Hapalogenys,
leiognathids, lethrinids, (lobotids?), pentacerotids, priacanthids
and Scoinbrops. To my knowledge, it occurs elsewhere only in
the larvae of holocentrid beryciforms and the caproid Antigonia.
The so called "tholichthys" larvae of the Chaetodontidae and
Scatophagidae (Fig. 262A-D) perhaps represent the ultimate in
head bone modification among larval percoids. The cranial bones
and many of the other exposed bones of the head are thickened
and rugose, effecting an armor-like protective covering. In chae-
todontids the posttemporal and supracleithrum are rugose and
expanded posteriorly as large laminar plates. The preopercle is
similarly expanded anteriorly and posteriorly and at its angle
bears a broad, flattened or serrated, terete spine. In scatophagids
the preopercle is rugose and expanded, but, unlike chaetodon-
tids, the supracleithrum is unmodified. The posttemporal is
rugose, its dorsal portion is somewhat expanded, and its ventral
half extends posteriorly as a very blunt, thick, spine-like pro-
jection. Also notable is a large, thick, rugose protuberance cov-
ering the pterotic. Although not identical, the larvae of chae-
todontids and scatophagids share a unique physiognomy, the
details of which should be investigated in relation to possible
close affinity of these two families.
Spination on circumorbital, nasal, premaxillary and maxillary
bones is generally found only in those larval percoids with cra-
nial ornamentation, and it is almost exclusively in these larvae
that other specializations, such as elongate serrate fin spines and
spinous scales occur. In addition, opercular and pectoral series
spination is usually more extensive and almost always includes
an elongate and/or serrate spine at the angle of the preopercle.
In summary, there seem to have been some common evo-
lutionary constraints on the order in which morphological com-
plexity and specialization of larval percoids has progressed, but
a simple direct relationship between this ordered progression
and phylogenetic affinity among families is not apparent. In fact,
the assemblages of taxa that characterize the various levels of
complexity discussed above are quite diverse and not compat-
ible with what little we do understand about percoid affinities
based on adult morphology. Furthermore, it is clear that elab-
orately ornamented larvae have arisen independently several
times within monophyletic groups otherwise characterized by
larvae with only generalized opercular and preopercular spi-
nation. Examples include the haemulid Conodon, the sparoid
family Lethrinidae and the serranid subfamily Anthiinae. Res-
olution of the phylogenetic significance of intricate patterns of
JOHNSON: PERCOIDEI
491
head spination among larval percoids will entail more precise
study than has characterized much previous work. Determi-
nation of homology will require detailed information about lo-
cation, conformation and processes of development of head
spines prior to considering the question of compatibility with
adult characters.
Utility of Larval Morphology in
Phylogenetic Studies
The preceding two decades have seen notable advances in our
understanding of the evolutionary relationships of teleost fishes;
however, as noted above, progress in elucidating the phylogeny
of the Percoidei has not kept pace. Many families are poorly
delineated and hypotheses about inter- and intrafamilial rela-
tionships are few. Lack of progress is chiefly attributable to the
size and diversity of the Percoidei. the adaptive malleability and
convergence that have characterized percoid evolution and the
paucity of conspicuous morphological specializations that can
be readily identified as true synapomorphies. With few excep-
tions (Burgess, 1974; Dooley, 1978; Kendall, 1979; Johnson,
1983), previous studies of percoid phylogeny and classification
have failed to consider early life history stages, even though it
is obvious that the prodigious variety of larval form and spe-
cialization among percoids offers a rich suite of additional char-
acters.
Within many families there is a complexity of larval mor-
phology or diversity of larval form that suggests excellent po-
tential for the application of larval characters in elucidating
generic interrelationships. Particularly promising families in this
regard include the Acropomatidae, Apogonidae, Bramidae, Ca-
rangidae, Cepolidae, Chaetodontidae, Haemulidae, Lutjanidae,
Malacanthidae, Pentacerotidae, Pomacanthidae, Priacanthidae,
Sciaenidae, and Serranidae. The intricate bony ornamentation
of the larvae of anthiin serranids, for instance, is considerably
more complex than that of the adults, and preliminary studies
of details of larval head spination and scale development among
New World genera indicate that the current generic classifica-
tion, based exclusively on adult morphology, should be reex-
amined (Carole Baldwin, Abstracts of 1 983 ASIH Annual Meet-
ing). Larvae of groups like the apogonids and carangids exhibit
a less complex morphology, but the wide range of form and
specialization should prove useful in phylogenetic analyses.
Larval morphology will undoubtedly also prove useful in con-
siderations of higher relationships among percoids. At the fam-
ily level, a rather simplistic approach is to consider that larvae
offer independent tests of hypotheses of monophyly. In other
words, do the larvae of each percoid family share one or more
derived features that corroborate the monophyly of that family
as currently defined on the basis of adult morphology? The
answer to this question appears to be yes for many groups, but
problems stem from an inadequate understanding of character
polarity and the fact that, for most families, larvae of many
genera and most species remain undescribed. Nonetheless, this
is a useful concept, and the validity and power of such a test
will increase as we gain more knowledge of the larvae of various
percoid groups.
Consider, for example, the bearing of larval morphology on
several hypotheses of relationship resulting from the recent re-
definition of Schultz's (1945) Emmelichthyidae, a polyphyletic
assemblage of planktivorous fishes. Heemstra and Randall (1977)
transferred Diptcrygonolus to the Caesionidae and Johnson
(1980) hypothesized that caesionids are lutjanoid fishes most
closely related to the lutjanid subfamily Lutjaninae. Caesionids
are quite distinctive in body form and upper jaw configuration,
but share with the lutjanines a number of osteological features
and a specialized adductor mandibulae (similar to that of most
carangids) in which a separate division of A, originates on the
subocular shelf Subsequent descriptions of larval lutjanines and
caesionids (see Table 122) show that they share a distinctive
body form, pattern of head spination, precocious first dorsal
and pelvic fins with elongate spines and soft rays, and sparse
pigmentation (Fig. 256A, B). The hypothesized sister group re-
lationship is thus corroborated by larval morphology.
The Centracanthidae were also removed from the Emme-
lichthyidae and hypothesized to be most closely related to the
Sparidae (Heemstra and Randall, 1977; Johnson, 1980) based
on adult morphology. Although the larvae of these two groups
share no obvious specializations, they are quite similar (Fig.
2581, J), and are distinguishable from those of the Emme-
lichthyidae (Fig. 2591) and the other reassigned groups. Labra-
coglossa, placed in a separate family by Heemstra and Randall
(1977) is here placed in the family Scorpididae (see section on
classification), and the larval form corroborates this placement
(Fig. 258A, B). The larvae of inermiids, Inermia and Emme-
lichthyops. also removed from the Emmelichthyidae, remain
undescribed, but their identification can provide a test of the
hypothesis that they are most closely related to the Haemulidae
(Johnson, 1980).
These examples and those that follow demonstrate that early
life history stages offer important information that can be used
to test previous phylogenetic hypotheses or incorporated with
adult characters into new phylogenetic analyses. Additional ex-
amples are mentioned in the discussion of familial classification.
Where the larvae are known, failure to consider their mor-
phology in studies of percoid phylogeny seems hardly justifiable,
and may inhibit progress or lead to false conclusions. This point
is well-illustrated in the two examples discussed below, in which
details of larval morphology provide critical evidence in support
of new or previously rejected phylogenetic hypotheses.
The families Branchiostegidae (=Latilidae) and Malacanthi-
dae have been variously united and separated in past classifi-
cations. In the most recent revision, Dooley (1978) concluded
that "the branchiostegids and malacanthids have few characters
in common that might be used to justify their consolidation
into a single family" and noted that they "could as easily be
aligned with several other percoid families as with each other."
He suggested that the malacanthids are possibly "a branch of
the labrid-scarid lineage, while the branchiostegids show closer
affinities to the serranid-percid line of perciform evolution." In
contrast, Robins et al. ( 1 980) recognized a close affinity between
the two groups by treating them as subfamilies of the Malacan-
thidae. Marino and Dooley (1982) took issue with this classi-
fication and stated that there are "several more myological (dif-
ferences) why the families are distinct." Actually, Marino and
Dooley listed only one myological difference, the absence of
adductor mandibulae section A,,,. This difference and the other
1 3 listed by Dooley ( 1 978. Table 1 ), including body depth, body
shape, and skull contour, have little relevance to the phyloge-
netic affinity of these two groups. As for features common to
the malacanthids and branchiostegids, Dooley found only three:
dorsal and anal fins relatively long and continuous, a single
opercular spine, and "grossly similar larval stages." Dooley cor-
rectly noted that the first two of these are not particularly mean-
ingful because they are fairly common percoid features, but he
492
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 122. References to Larval Percoidei.
Eggs
Poslflexion
Acanthoclinidae
Acropomatidae
Ambassidae
Apogonidae
Bramidae
Caesionidae
Callanthiidae
Carangidae
Caristiidae
Centracanthidae
Centrarchidae
Centropomidae
Cepolidae
Chaetodontidae
Cheilodactylidae
Cirrhitidae
Congrogadidae
Coracinidae
Coryphaenidae
Jillett, 1968
Breder and Rosen, 1966
Eng, 1969
Nair, 1958
Breder and Rosen, 1966
Leis and Rennis, 1983
Allen, 1975b
Bertolini, 1933a
Jillett, 1968
Eng, 1969
Jillett, 1968
Leis and Rennis, 1983
Miller etal., 1979
Allen, 1975b
Bertolini, 1933a
Leis and Rennis, 1983
Miller etal., 1979
Allen, 1975b
De Gaetani, 1937
Johnson, 1978
Mead, 1972
Johnson, 1978
Mead, 1972
— — Leis and Rennis, 1983
Bertolini, 1933b
Page, 1918
Laroche et a!., this volume
Brownell, 1979
Thomopoulos, 1954
Aboussouan, 1964
Montalenti, 1933
Sanzo, 1939c
Numerous references, see Breder and Rosen, 1966; Hardy, 1978b; and Auer, 1982
Brownell, 1979
Sanzo, 1939c
Brownell, 1979
Sanzo, 1939c
Lauand Shafland, 1982
Breder and Rosen, 1966
Russell, 1976
Holt, 1891
Montalenti, 1937b
Leis and Rennis, 1983
Burgess, 1978
Suzuki etal., 1980
Brownell, 1979
Mito, 1963
Robertson, 1978
Gilchrist and Hunter,
1919
Barnard, 1927
Lauand Shafland, 1982
Leis and Rennis, 1983
Suzuki etal., 1980
Brownell, 1979
Robertson, 1978
Johnson, 1978
Miller etal., 1979
Mito, 1960
Johnson, 1978
Miller etal., 1979
Mito, 1960
Lau and Shafland, 1982
Russell, 1976
Page, 1918
Montalenti, 1937b
Okiyama, 1982b
Leis and Rennis, 1983
Suzuki etal., 1980
Brownell, 1979
Gilchnst and Hunter, 1919
Hatton, 1964
Leis and Rennis, 1983
Johnson, 1978
Miller etal., 1979
Mito, 1960
Potthoff, 1980
Crossland, 1981
Crossland, 1982
Jillett, 1968
Pourmanoir, 1976
Okiyama, 1982b
Nair, 1952b
Gopinath, 1946
Nair, 1958
Leis and Rennis, 1983
Miller etal., 1979
Allen, 1975b
Pourmanoir, 1976
Okiyama, 1982b
Bertolini, 1933a
Pahay, 1975
Whitley, 1926
Vatanachi, 1972
De Gaetani, 1937
Johnson, 1978
Mead, 1972
Pahay, 1983
Leis and Rennis, 1983
Leis and Rennis, 1983
Pourmanoir, 1976
Bertolini, 1933b
Page. 1918
Belyanina, 1982b
Brownell, 1979
Page, 1918
Montalenti, 1933
Lau and Shafland, 1982
Russell, 1976
Pourmanoir, 1976
Clark, 1920
Page, 1918
Montalenti, 1937
Pourmanoir, 1973
Leis and Rennis, 1983
Burgess, 1978
Pourmanoir, 1976
Kendall and Goldsborough,
1911
Burgess, 1974
Brownell, 1979
Dudnik, 1977
Vooren, 1972
Tong and Saito, 1977
Nielsen, 1963a
Hatton, 1964
Leis and Rennis, 1983
Pourmanoir, 1973
Pourmanoir, 1971a
Whitley, 1926
Smith, 1938
Johnson, 1978
Miller et al., 1979
Gibbs and Collette. 1959
Aboussouan. 1969
Potthotf, 1980
JOHNSON: PERCOIDEI
493
Table 122. Continued.
Eggs
Yolk-sac
Preflcxion
Poslflexion
Echeneididae
Emmelichthyidae
Ephippididae
Epigonrdae
Gerreidae
Girellidae
Haemulidae
Hapalogenys
Howella
Kyphosidae
Lactariidae
Laleolabrax
Leiognathidae
Lethrinidae
Lobotidae
Lutjanidae
John, 1950
Sanzo, 1930a
Martin and Drewry, 1978
Sanzo. 1928
Akazaki et al., 1976
Breder and Rosen 1966
Johnson. 1978
Ryder. 1887
Leis and Rennis, 1983
Rass. 1972
Breder and Rosen, 1966
Uchidaelal., 1958
Mito, 1957a
Breder and Rosen, 1966
Leis and Rennis. 1983
Johnson. 1978
Mito. 1966
Podosinnikov. 1977
Saksena and Richards.
1975
Hildebrand and Cable.
1930
Fahay. 1983
Suzuki et al.. 1983
Leis and Rennis. 1983
Miller et al.. 1979
Watson and Leis. 1974
Breder and Rosen. 1966
Chacko. 1944
Breder and Rosen. 1966
Mito, 1957b
Uchidaet al., 1958
Breder and Rosen, 1966
Fujita, 1960
Leis and Rennis, 1983
Suzuki and Hioki, 1978
Renzhai and Suif'en,
1980a
Mito, 1956a
Hardy, 1978b
Gudger, 1931
Leis and Rennis, 1983
Suzuki and Hioki, 1979b
Rabalaiset al., 1980
Stark, 1971
Mon, 1984
John, 1950
Sanzo, 1930a
Martin and Drewry. 1978
Sanzo, 1928
Akazaki et al.. 1976
Johnson, 1978
Ryder, 1887
Leis and Rennis, 1983
Uchidaet al., 1958
Mito, 1957a
Leis and Rennis, 1983
Johnson, 1978
Mito, 1966
Podosinnikov, 1977
Saksena and Richards,
1975
Hildebrand and Cable,
1930
Fahay. 1983
Suzuki et al., 1983
Leis and Rennis, 1983
Miller etal.. 1979
Mito. 1957b
Uchida et al..
Fujita. 1960
1958
Leis and Rennis. 1983
Suzuki and Hioki. 1978
Renzhai and Suifen.
1980a
Mito. 1956a
Leis and Rennis. 1983
Suzuki and Hioki. 1979b
Rabalais et al., 1980
Mori, 1984
John. 1950
Martin and Drewry, 1978
Sanzo. 1928
Akazaki et al.. 1976
Johnson. 1978
Hildebrand and Cable,
1938
Fahay, 1983
Leis and Rennis, 1983
Uchidaet al., 1958
Mho, 1957a
Leis and Rennis, 1983
Johnson, 1978
Saksena and Richards,
1975
Hildebrand and Cable,
1930
Fahay, 1983
Suzuki et al., 1983
Gonzales. 1946
Leis and Rennis, 1983
Miller etal., 1979
Mito, 1957b
Uchidaet al., 1958
Malacanthidae
Breder and Rosen, 1966
Fischer. 1958
Fischer, 1958a
Fahay, 1983
Fujita, 1960
Leis and Rennis, 1983
Hardy, 1978b
Uchidaet al., 1958
Leis and Rennis, 1983
Richards and Saksena.
1980
Collins et al.. 1980
Laroche. 1977
Mon. 1984
Fischer. 1958a
Okiyama, 1964
Gudger, 1926
Gudger, 1928
Akazaki et al., 1976
Nakahara, 1962
Johnson, 1978
Hildebrand and Cable. 1938
Fahay. 1983
Mayer. 1972
Leis and Rennis. 1983
Nair, 1952b
Uchidaet al., 1958
Kobayashi and Igarashi,
1961
Munro, 1945
Uchidaet al., 1958
Leis and Rennis, 1983
Johnson, 1978
Saksena and Richards, 1975
Hildebrand and Cable, 1930
Nellen, 1973b
Fahay, 1983
Heemstra, 1974
Okiyama, 1982b
Suzuki etal., 1983
Gonzales, 1946
Leis and Rennis, 1983
Moore, 1962
Johnson, 1978
Uchidaet al., 1958
Nair, 1952b
Okiyama, 1982b
Mito, 1957b
Uchida etal., 1958
Nair, 1952b
Vatanachi, 1972
Gopinath, 1946
Leis and Rennis, 1983
Hardy, 1978b
Okiyama, 1982b
Uchidaet al., 1958
Leis and Rennis, 1983
Fourmanoir, 1976
Okiyama, 1982b
Richards and Saksena, 1980
Collins et al., 1980
Fahay, 1975
Heemstra, 1974
Vatanachi, 1972
Stark, 1971
Musiy and Sergiyenko, 1977
Laroche, 1977; Mori, 1984
Fourmanoir, 1970, 1976
Dooley, 1978
494
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 122. Continued.
Eggs
Preflexion
Fahay, 1983
Microcanthidae
Monodactylidae
Moronidae
Mullidae
Nemipteridae
Opistognathidae
Oplegnathidae
Pempheridae
Pentacerotidae
Percichthyidae
Percidae
Plesiopodae
Polyprion
Pomacanthidae
Pomatomidae
Akatsuet al., 1977
Breder and Rosen. 1966
Hardy, 1978b
Mansueti, 1964
Ryder, 1887
Mansueti, 1958
Pearson, 1938
Breder and Rosen, 1966
Leis and Rennis, 1983
Russell, 1976
Miller et al., 1979
Marinaro, 1971
Raffaele, 1888
Heincke and Ehrenbaum,
1900
Leis and Rennis, 1983
Aoyama and Sotogaki,
1955
Renzhai and Suifen,
1980b
Breder and Rosen, 1966
Mito, 1956b
Uchidaet al., 1958
Leis and Rennis, 1983
Breder and Rosen, 1 966
Dakin and Kesteven,
1938
Llewellyn, 1974
Lake, 1967
Jackson, 1978
Fuster de Plaza and Plaza,
1955
Akatsuet al., 1977
Hardy, 1978b
Mansueti, 1964
Ryder, 1887
Mansueti, 1958
Pearson, 1938
Doroshev, 1970
Leis and Rennis. 1983
Russell, 1976
Marinaro, 1971
Raffaele, 1888
Heincke and Ehrenbaum,
1900
Leis and Rennis, 1983
Aoyama and Sotogaki,
1955
Renzhai and Suifen,
1980b
Fukuharaand Ito, 1978
Mito, 1956b
Uchidaet al., 1958
Leis and Rennis, 1983
Dakin and Kesteven,
1938
Llewellyn, 1974
Lake. 1967
Jackson, 1978
Fahay, 1983
Leis and Rennis, 1983
Uchidaet al., 1958
Akatsuet al., 1977
Hardy, 1978b
Mansueti, 1964
Ryder, 1887
Mansueti, 1958
Pearson, 1938
Doroshev, 1970
Fritzsche and Johnson,
1980
Leis and Rennis, 1983
Russell, 1976
Miller etal., 1979
Heincke and Ehrenbaum,
1900
Montalenti, 1937
Uchidaet al., 1958
Lo Bianco, 1908b
Leis and Rennis, 1983
Fukuhara and Ito, 1978
Uchidaet al., 1958
Leis and Rennis, 1983
Dakin and Kesteven, 1938
Llewellyn, 1974
Lake, 1967
Jackson, 1978
Numerous references, see Breder and Rosen, 1966; Hardy, 1978b; and Auer, 1982
Breder and Rosen, 1 966
Mito, 1955
Hardy, 1978b
Sparta, 1939a
Thomson and Anderton,
1921
Leis and Rennis, 1983
Suzuki etal.. 1979
Fujita and Mito, 1960
Hardy, 1978b
Deuel etal., 1966
Dekhnik, 1973
Salekhova, 1959
Sparta, 1962
Fahay, 1983
Mito, 1955
Hardy, 1978b
Sparta, 1939a
Leis and Rennis, 1983
Suzuki etal., 1979
Fujita and Mito, 1960
Hardy, 1978b
Deuel et al., 1966
Dikhnik, 1973
Salekhova, 1959
Sparta, 1962
Fahay, 1983
Hardy, 1978b
Sparta, 1939a
Leis and Rennis, 1983
Burgess, 1974
Hardy, 1978b
Deuei et al., 1966
Dekhnik, 1973
Salekhova, 1959
Sparta, 1962
Norcross et al., 1974
Pearson, 1941
Fahay, 1983
Moser, 1981
Okiyama, 1964
Okiyama, 1982b
Fahay, 1983
Berry, 1958
Hubbs, 1958
Leis and Rennis, 1983
Uchidaetal., 1958
Akatsu et al., 1977
Ogasawara et al., 1978
Hardy, 1978b
Mansueti, 1964
Mansueti, 1958
Pearson, 1938
Doroshev, 1970
Okiyama, 1982b
Fritzsche and Johnson, 1980
Leis and Rennis, 1983
Johnson, 1978
Russell. 1976
Miller etal., 1979
Uchidaetal., 1958
Vatanachi, 1972
M. C. Caldwell, 1962
Lo Bianco, 1908b
Leis and Rennis, 1983
Vatanachi, 1972
Fukuharaand Ito, 1978
Fuskusho, 1975
Leis and Rennis, 1983
Zama el al.. 1977
Hardy, 1982
Dakin and Kesteven, 1938
Lake, 1967
Jackson, 1978
Hardy, 1978b
Sparta, 1939a
Bertolini, 1933b
Leis and Rennis, 1983
Burgess, 1978
Fourmanoir, 1976
Burgess, 1974
Hardy, 1978b
Dekhnik, 1973
Salekhova, 1959
Norcross et al., 1974
Pearson, 1941
Fahay, 1983
Silverman, 1975
JOHNSON: PERCOIDEI
495
Table 122. Continued.
Eggs
Prcficxion
Postflexion
Priacanthidae
Pseudochromidae
Rachycentridae
Scatophagidae
Sciaenidae
Scorpididae
Serranidae
Sillaginidae
Siniperca
Sparidae
Stereolepis
Symphysanodon
Terapondiae
Leis and Rennis, 1983
Suzuki et al., 1980
Leis and Rennis, 1983
Lubbock. 1975
Hardy, 1978b
Leis and Rennis,
Lubbock, 1975
1983
Leis and Rennis, 1983
Hardy, 1978b
D. K. Caldwell, 1962
Aboussouan, 1969
Leis and Rennis, 1983
— Weber and de Beaufort, —
1936
Numerous references, see Breder and Rosen, 1966; Hardy, 1978b; and Auer, 1982
- - Hattori. 1964
Kendall, this volume
Breder and Rosen, 1966
Ueno and Fujita, 1954
Uchidaet al.. 1958
Ueno and Fujita, 1954
Uchidaet al., 1958
Munro, 1945
Uchidaet al.. 1958
Imai and Nakahara, 1957
Chyung, 1977
Breder and Rosen, 1966
Johnson, 1978
Russell, 1976
Ranzi, 1933
Rathbun, 1893
Cardeilhac, 1976
Kuntzand Radcliffe, 1917
Houde and Potthoff, 1976
Uchidaet al., 1958
Fahay, 1983
Hussain et al., 1981
Breder and Rosen, 1966
Llewellyn, 1973
Zvjagina, 1965b
Lake, 1967
Imai and Nakahara,
Chyung, 1977
1957
Imai and Nakahara,
Chyung, 1977
1957
Johnson, 1978
Russell, 1976
Ranzi, 1933
Kuntzand Radcliffe, 1917
Houde and Potthoff, 1976
Uchidaet al., 1958
Fahay, 1983
Kohnoet al.. 1983
Hussain et al., 1981
Llewellyn, 1973
Uke. 1967
Johnson, 1978
Russell. 1976
Ranzi, 1933
Hildebrand and Cable,
1930
Kuntzand Radcliffe, 1917
Houde and Potthoff, 1976
Fahay. 1983
Kohnoet al.. 1983
Hussam et al., 1981
Llewellyn, 1973
Zvjagina, 1965b
Uke, 1967
Leis and Rennis, 1983
Hardy, 1978b
D. K. Caldwell, 1962
Fourmanoir, 1976
Okiyama, 1982b
Leis and Rennis, 1983
Hardy, 1978b
Dawson, 1971a
Nair, 1952b
Weber and de Beaufort, 1936
Hattori, 1964
Okiyama, 1982b
Munro, 1945
Uchidaet al., 1958
Gopinath. 1946
Okiyama, 1982b
Imai and Nakahara, 1957
Chyung, 1977
Johnson, 1978
Russell, 1976
Ranzi, 1933
Hildebrand and Cable, 1930
Kuntz and Radcliffe, 1917
Okiyama, 1982b
Munro, 1945
Houde and Potthoff, 1976
Uchidaet al., 1958
Fahay, 1983
Kohnoet al., 1983
Hussain et al., 1981
Okiyama, 1982b
Fourmanoir, 1973
Llewellyn, 1973
Nair, 1952b
Munro. 1945
Zvjagina. 1965b
Lake, 1967
Vatanachi, 1972
incorrectly dismissed the significance of the larvae, which, as
Okiyama (1982b) pointed out, are remarkably similar and dis-
tinctive among the percoids. I believe the larval morphology of
these two groups offers conclusive evidence for a sister-group
relationship between them, including a synapomorphy unique
among percoids, and perhaps all teleosts.
Larval malacanthids and branchiostegids (Fig. 256E, ¥), are
among the most elaborately ornamented in the Percoidei. They
share early developing spinous scales, a series of serrate ridges
on the frontals, and have very similar configurations of spines
and serrate ridges on many of the exposed bones of the head.
The most distinctive feature is a median rostral bony structure,
forming a blunt, serrate-ridged projection in Caulolatilus. Lo-
pholattlus and Branchiostegus. a smooth anchor-shaped projec-
tion in Malacanthus and a long spike-like spine with serrate
ridges in Hoplolatilus. Dooley (1978) stated that larvae with
similar rostra and head spination occur among holocentrids,
lutjanids, serranids and istiophorids and thai the similarity "could
be considered as convergence or perhaps a relict characteristic
carried over from a common beryciform ancestor." In fact, the
larvae of these groups are quite different morphologically, and
misconceptions about their similarity apparently result from
superficial considerations that have often characterized earlier
larval descriptions. Neither larval lutjanids nor serranids have
rostral projections or (with the exception of some anthiin ser-
ranids) particularly elaborate head spination. The rostral pro-
jection of istiophorids is a premaxillary beak or bill, supported
internally by a fixed, horizontally-oriented rostral cartilage and
is structurally homologous to that of larval Xiphias and scom-
brids (except Scombrini). Although the spinous rostrum of hol-
ocentrids bears a strong resemblance to that of Hoplolatilus. it
is an entirely different structure, formed by enlargement of the
supraethmoid and supported by a greatly enlarged ethmoid car-
tilage. The median rostral projection of malacanthids and bran-
chiostegids has been described as an ethmoid spine (Okiyama,
1964, 1982b), but it actually originates from a modification of
496
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 263. Scanning electromicrographs of epithelium of juvenile dolphins and cobia at various magnifications. (A) Coryphaena hippurus, 28
mm SL, 15 x; (B) C. hippurus. 28 mm SL, 360 x; (C) Rachycentron canadum. 30 mm SL, 15 x; and (D) R. canadum. 80 mm SL, 360 x.
the nasal bones. The nasal bones first appear as separate struc-
tures, but prior to or during flexion, they become fused anteriorly
by a median bony bridge. This modified nasal structure then
develops the various ornamentations that characterize mala-
canthid and branchiostegid larvae. At transformation, the bony
bridge begins to fragment and is eventually entirely resorbed,
so that the nasal bones once again become completely separate.
I know of no other example in fishes of transient ontogenetic
fusion of nasal bones. This unique synapomorphy, in conjunc-
tion with the other shared larval specializations, cogently sup-
ports the hypothesis that malacanthids and branchiostegids are
sister groups. Classification of the two lineages of tilefishes as
subfamilies of the Malacanthidae seems an appropriate way to
express this relationship.
The evolutionary relationships of the dolphins, Coryphaen-
idae, have remained uncertain, but the family has usually been
placed close to the Carangidae as have the Echeneididae and
the monotypic Rachycentridae. Examination of the larvae of
these groups during this investigation and subsequent consid-
erations of adult morphology have led to further resolution of
the interrelationships of these families (Johnson, Abstracts of
1983 ASIH Annual Meeting). This final example provides the
most convincing illustration of the importance of larval char-
acters to studies of phylogeny among percoids. Consequently I
discuss it in considerable detail.
Freihofer (1978) noted that the Nematistiidae, Carangidae,
Coryphaenidae, Rachycentridae and Echeneididae share a unique
specialization in the lateralis system on the snout— an anterior
extension of the nasal canal consisting of one (Nematistiidae)
or two prenasal canal units, with one (Nematistiidae and Ca-
rangidae) or both (remaining three families) surrounded by tu-
bular ossifications. In addition, they share small, adherent cy-
cloid scales. Based on two presumed synapomorphies, then,
these five families constitute a monophyletic group, hereafter
referred to as the carangoids.
Three synapomorphies unite the Carangidae, Coryphaenidae,
Rachycentridae and Echeneididae as a monophyletic group.
These four families lack the bony stay (Potthoff, 1975) posterior
to the ultimate dorsal and anal pterygiophores found in almost
all other percoids (see Table 1 20), have two prenasal canal units
and have a lamellar expansion along the anterior margin of the
coracoid. Nematisttus, placed in separate family by Rosenblatt
and Bell (1976), is apparently the sister group of these four
families (see cladogram. Fig. 276, in Smith-Vaniz, this volume).
JOHNSON: PERCOIDEI
497
Fig. 264. Scjiiaiiig clcctromicrographs of epillicliuiu ol laival Uolpliin and amberjack at various magnifications. (A) Coryphaena hippurus,
17.0 mm SL, 55x;(B) C. hippurus. 17.0 mm SL, 400x;(C) Seriola sp., 11.2 mm SL, 55x;and(D) 5. sp., 11.2 mm SL, 2,000 x.
It has a well developed bony stay, a single, partly ossified pre-
nasal canal unit and an unmodified coracoid.
Within the carangoids, the Coryphaenidae, Rachycentridae
and Echeneididae form a monophyletic group, here referred to
as the echeneoids. Adult echeneoids are specialized with respect
to the Carangidae in the following features: absence of predorsal
bones; anterior shift of the first dorsal pterygiophore forward of
the third intemeural space; presence of several anal pterygio-
phores anterior to the first haemal spine (vs. one in carangids
and most other percoids); loss of the so-called beryciform fo-
ramen in the anterior ceratohyal; and tubular ossifications sur-
rounding both prenasal canal units. Larval echeneoids are also
specialized with respect to carangids (larvae of Ncmatistius are
unknown). Whereas larval carangids are moderate to deep-bod-
ied, hatch at small sizes (1-3.5 mm) and complete dorsal fin
and anal fin rays in conjunction with or soon after flexion,
echeneoid larvae (Fig. 261 A-C) are very elongate, hatch at large
sizes and complete dorsal fin rays at two to three times the size
at flexion (sec Table 121). Larval morphology thereby corrob-
orates the hypothesized monophyly of the echeneoids.
Although a sister-group relationship between the Coryphaen-
idae and either the Rachycentridae or the Echeneididae has not
been previously proposed, it has often been suggested that
Rachycentron and the echeneidids are sister groups. This hy-
pothesis was based on general external similarity including the
remarkable resemblance in body form, color pattern and caudal
fin shape between juveniles of Rachycentron and Echeneis nau-
aa?«(B6hlke and Chaplin, 1968). Because the juvenile features
of Rachycentron are shared by only one species of echeneidid,
they do not provide evidence for a sister-group relationship
between the Rachycentridae and the Echeneididae, nor does a
detailed osteological comparison of the two groups. The eche-
neidids are highly modified in almost every aspect of their os-
teology compared to both Rachycentron and Coryphaena, and
with two exceptions (absence of a median cranial crest and
fusion of the prenasal ossifications), the only specializations
shared by both Rachycentron and the echeneidids are also shared
by Coryphaena. The following are autapomorphies of the Eche-
neididae: spinous dorsal fin modified as an attachment disc
covenng the dorsal surface of the cranium; first neural arch fused
498
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
to its centrum, spine absent; endopterygoid absent; quadrate
with a lateral shelf; palatine and upper jaw bones distinctively
modified; postcleithra absent; supracleithrum extremely re-
duced; medial tabular bones absent; posttemporal modified in
shape and angle of articulation with supracleithrum; pelvic gir-
dle broad and short, with two distinct anterior processes; caudal
skeleton with a full neural spine on the second preural centrum;
branchial skeleton with main arm of first epibranchial reduced
to a nubbin, uncinate process enlarged and articulating directly
with second pharyngobranchial, and interarcual cartilage absent.
None of these extreme modifications (those of the caudal and
branchial skeletons being unique among percoids) are even fore-
shadowed in the skeleton of Rachycentron, which is instead
remarkably similar to that oi Coryphaena. except in the anterior
portion of the dorsal fin and the neurocranium.
In Coryphaena, the dorsal fin is elaborated anteriorly and
extended into the first intemeural space (second in Rachycen-
tron) and there is an extreme supraoccipito-frontal crest on the
neurocranium. The dorsal fin modification is autapomorphic
for Coryphaena. but the median cranial crest is probably prim-
itive for echeneoids since it is variously developed in all caran-
gids and well-developed in Nematislius. The absence of this
crest in Rachycentron, associated with a slight flattening of the
neurocranium, is the only specialization shared with the Eche-
neididae. Here again, however, there is little similarity between
the slightly flattened neurocranium of Rachycentron and the
extremely flattened and restructured neurocranium of the eche-
neidids, in which, for instance, the supraethmoid and vomer
have become flat plates and the orbit is completely occluded by
enlargement and anterior extension of the pterosphenoids. This
extreme restructuring of most cranial bones is evident even in
larval echeneidids at the earliest development of the neuro-
cranium, whereas the neurocrania of Rachycentron and Cory-
phaena exhibit a generalized development similar to that of
carangids. Prior to development of the median crest in Cory-
phaena(> 100 mm), the neurocrania of cobia and dolphin differ
mainly in relative depth. Echeneidids also have an exceptionally
modified adductor mandibulae in which A, is absent and A,
and A„ are distinctively subdivided. Coryphaena and Rachy-
centron share a relatively generalized adductor mandibulae, spe-
cialized with respect to the primitive carangids (see section on
Carangidae) in having A, somewhat reduced and inserting nar-
rowly on the maxillo-mandibular ligament.
The pronounced similarities between Coryphaena and Rachy-
centron in the adductor manidbulae and most osteological fea-
tures merely serve to reiterate the lack of evidence for the fre-
quently proposed sister-group relationship between Rachycentron
and the echeneidids. Further comparison with character states
throughout the Carangidae will be required to define these adult
similarities as primitive or derived features. The most com-
pelling evidence for a sister-group relationship between Cory-
phaena and Rachycentron is found in the morphology of their
larvae. As noted above all echeneoid larvae have a similar body
form and pattern of development, but the elongate, flattened
head of larval echeneidids lacks ornamentation. In contrast,
larval dolphin and cobia share identical patterns of head spi-
nation: a small posttemporal spine; several spines on the pos-
terior and lateral margin of the preopercle, including one en-
larged spine on either side of its angle; and a very large,
posterolaterally directed spine on the supraorbital ridge of each
frontal bone. Another obvious feature is the presence of laterally
swollen pterotics, previously described in Coryphaena as blunt
sphenotic spines (Gibbs and Collette, 1959). This specific pat-
tern of head spines is distinctive, but similar features occur in
various combinations among carangid larvae, and it is pre-
mature to interpret this configuration as synapomorphic for
Coryphaena and Rachycentron until detailed comparisons with
carangids have been made.
A specialization clearly unique to the larvae of dolphin and
cobia, however, is a modified epithelial cuticle in which are
borne minute crown-shaped spicules (Figs. 263A-D, 264A, B).
The cuticle itself is composed of large, multinucleate "cells,"
40-100 m in diameter, that appear to continually produce and
slough-off" the thorny spicules. Each epithelial "cell" produces
one spicule, so that these extraordinary structures cover all ex-
posed body surfaces, excluding the pupil of the eye, giving the
integument a bristly appearance under magnification (Fig. 264A).
They first appear at about 8mm and are present in some indi-
viduals as large as 100 mm. Further histological work and elec-
tron microscopy will be necessary to determine the composition
of the spicules, which may be keratinous. It is clear, however,
that they are neither bony nor cartilaginous. Their function is
unknown, but as with spinous scales, it seems likely that they
are defensive.
The surface and cellular composition of the epithelium of
larval echeneidids appear normal, but some modification of the
larval epithelium may actually be a primitive feature of car-
angoids. In larvae of trachinotine and naucratine carangids ex-
amined thus far (Trachinotus, Naucrates, Seriola) the epithelial
cells are of normal size ( ~ 8- 1 2 m), but their surfaces bear clusters
of bumplike structures, seemingly the result of keratinization
(Fig. 264C, D). Absence of these modified epithelial cells in
larvae of carangine carangids is parsimoniously interpreted as
secondary (see Laroche et al., this volume). Their presence in
the larvae of Neinalistius (curtently unknown) would corrobo-
rate the hypothesis that modified larval epithelium is primitive
for carangoids and thus also for echeneioids, suggesting that it
has been lost in carangines and echeneidids.
The multinucleate epithelial cells and enlarged, thorny spic-
ules of larval Coryphaena and Rachycentron represent a com-
plex, shared specialization, unique among percoids. The phy-
logenetic significance of this synapomorphy is lessened only by
the unlikely possibility that loss of a modified epithelium in
echeneidids occurted after development of multinucleate cells
and spicules. Available evidence strongly points to a Cory-
phaena-Rachycentron sister-group relationship, and it should
be clear that further investigations testing this hypothesis must
integrate larval, adult and developmental characters.
In conclusion, the study of early life history stages of fishes
has traditionally been treated as a discipline somewhat removed
from the mainstream of systematic ichthyology. As a result,
larval morphology has rarely beeen incorporated into studies of
evolutionary relationships of fishes. It is evident that the larvae
of percoid fishes exhibit a prodigious array of complexity and
diversity that offers exceptional potential applicability to phy-
logenetic studies. Recognition and application of this potential
will be an important step in understanding the complex evo-
lutionary history of the Percoidei.
South Carolina Wildlife and Marine Resources De-
partment, Post Office Box 12559, Charleston, South
Carolina 29412. Present Address: Fish Division, Na-
tional Museum of Natural History, Washington,
District of Columbia 20560.
Serranidae: Development and Relationships
A. W. Kendall, jr.
THE percoid family Serranidae is defined by the presence of
three spines on the opercle (Goshne, 1966) and three re-
ductive specializations (absence of the posterior uroneural, pro-
current spur, and third preural radial cartilage) that separate it
from the Percichthyidae (Johnson, 1983). These are primarily
tropical to temperate marine fishes that vary in size from < 10
cm to >300 cm. It is a speciose family with nearly 400 species
(Nelson, 1976) that has had a history of being hard to charac-
terize and subdivide. The serranids are continuing objects of
taxonomic studies from the species to subfamily levels and sev-
eral new species are described each year, primarily anthiines
whose deep-water reef habitat has made collecting difficult. As
presently understood (Johnson, 1983). the family is composed
of 3 subfamilies (Serraninae. Anthiinae. and Epinephelinae),
although Katayama (1960) recognized 15 subfamilies. Various
authors have included other groups (e.g.. Callanthias) in the
Serranidae, and others have raised parts of the family to familial
status (e.g., Anthiinae and Grammistinae). Such problems will
probably not be resolved without a worldwide revision of the
family, which is not forthcoming.
Development
The eggs of all but a few serranids are unknown. Often, Wil-
son's (1891) classic work on the development of Centropnsiis
striata eggs has been cited as the example of teleost embrjology
in texts (e.g., Nelsen, 1953). Serranid eggs described to date are
typical of the majority of pelagic marine teleost eggs: they are
spherical, about 1 mm in diameter, have a single oil globule, a
narrow perivitelline space, and a smooth egg envelope. Several
species of Epincphelus (e.g., Guitart Manday and Juarez Fer-
nandez, 1966: Hussain and Higuchi, 1980), fara/aira.v (Butler
et al., 1982), and several anthiines (e.g.. Suzuki et al., 1974,
1978) have been reared. There seems to be a difference in oil
globule placement in yolk-sac larvae among the subfamilies (Fig.
265). Larvae of representatives of all the subfamilies, most of
the tribes, and about a third of the genera of serranids have
been described. Serranid larvae fall into one of four types, which
correspond to two of the subfamilies and two of the tribes within
the Epinephelinae. These larval types can be characterized based
on the taxa for which larvae are known as follows (based on
Kendall, 1979).
Serraninae. — hody proportions show rather direct develop-
ment. There are no elongate spines in the opercular region,
rather a series of blunt points. The fin spines are thin and only
slightly elongated in some. Most larval pigment consists of me-
lanophores in characteristic positions along the ventral midline.
Anthiinae.— These deep-bodied larvae have produced spines on
several bones in the opercular region, some of which may be
serrated. There is a tendency to develop armature on the head,
and the interopercular has a characteristic long posteriorly di-
rected spine that is overlaid by an even larger, similar spine on
the preopercular. The pelvic and some dorsal fin spines are
strong, serrate in some, and not very elongate. Pigment consists
mainly of large blotches and dashes in characteristic positions
on the trunk.
Epinephelini. — Knovfn larvae of members of this tribe are all
quite similar and generally difficult to assign to a genus on the
basis of larval characters. These are among the most spectacular
offish larvae, with stout, elongate, serrate, and pigmented dorsal
and pelvic fin spines. Usually the second dorsal spine is much
longer than the others and it, as well as the pelvic spines, are
as long as the body. The dorsal spine is often "locked" in an
upright position— presumably possible because of a unique pte-
Fig. 265. Newly hatched yolk-sac larvae of serranids: (A) Serraninae:
Paralabrax clathratus. from Butler et al. (1982); (B) Anthiinae: Sacura
marganlacea. from Suzuki et al. (1974); and (C) Epinephelinae: Epi-
nephelus akaara. from Ukawa et al. (1966).
499
500
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 123. Serranid Taxa (Subfamilies through Subgenera). Their general distribution and references to early life history (ELH) descriptions.
A-A: Atlanto-American. I-P; Indopacific. Stages described: E (eggs), Y (yolk-sac larvae), L (larvae — yolk-sac through post flexion), Pr (preflexion
larvae), F (flexion larvae), Po (postflexion larvae), T (transforming larvae) and J (juveniles).
Subfamily
Genus
subgenus
Distn
bution
Iribe
A-a
IP
Serraninae
Acanlhislius
+
Cenlropnstis
+
Chelidoperca
+
Crattnus
+
Diplectrum
-1-
Dules
-1-
Hypoplectrus
+
Paratabrax
+
Schultzea
+
Senamculus
-1-
Serramis
+
Serranus
Paracenlropnstis
Anthiinae
Anthias
Microlahrichthys
Nemanthias
Pseudanlhias
+
+
Caesioperca
+
Caprodon
-1-
-1-
Daciylanthias
+
Ellerketdia
-t-
Franzia
+
Giganthias
+
Hemanlhias
+
Holanthias
+
+
Luzonichlhys
+
Ocyanthias
+
Odontanlhias
+
Pleclranthias^
-1-
+
Pronotogrammus
+
Sacura
+
Selenanthias
+
Senanocirrhitus
+
Tosana
+
Tosanoides
+
Epinephelinae''
Niphonini
Niphon
+
Epinephelini
A nyperodon
+
Cromileplcs
+
Epmephelus^
-1-
+
ELH descnplions
Alphestes
Cephalopholis
Dermalolepis
Epinephelus
Promicrops
Gonioplectrus
+
Gracilia
+
Mvcleroperccf
+
Paranthias
+
+
Pleclropomus
-1-
Trisolropis
+
Variola
+
Diploprionini
Aulacocephalus
+
Belonoperca
+
Diploprion
+
Liopropomini
Jehoehlkia
-1-
Liopropoma
+
+
Ryder (1888)-E-Y, Wilson (189I)-E-Y, Hofr(1970)-E-Y, Kendall (1972)-Pr-J.
Kendall (1977, l979)-Pr-Po
Kendall (1977, l979)-Pr-Po
Kendall (1977, l979)-Pr-Po
Kendall (1977, 1979)-Pr-Po, Butler et al. (1982)-E-J
Kendall (1977, l979)-Pr-Po
Rafl!"aele (1888)-E,Y, Page (1918)-Pr-J, Roule and Angel (I930)-L, Bertolini
(1933b)-E-J, Vodyanitsky and Kazanova (1954)-E-Po, Aboussouan (1972b)-Pr-
Po, Kendall (1977, 1979)-Pr-Po
Roule and Angel (I930)-Pr-T, Sparta (1932)- Pr-T, Bertolini (1933b)-Pr-J, Abous-
souan (1972b)-Pr-Po, Fourmanoir (l976)-Po, Kendall (1977, 1979)-Pr-Po, Su-
zuki et al. (in press), Leis and Rennis (1983)— Pr-T
Fourmanoir (1976)— Po
Suzuki el al. (1978)-E-Y
Kendall (1977, 1979)-Pr-Po
Fourmanoir (1976) — T
Kendall (1977, 1979)-Po
Kendall (1977, l979)-Pr-Po
Suzuki et al. (I974)-E-Pr, Fourmanoir (1976)- Po
Fourmanoir (1973) — T
Raffaele (I888)-E, Page (19l8)-Po, Bertolini (1933b)-F, Sparta (1935)-E-T, Fow-
ler (1944)— T, Vodyanitsky and Kazanova (1954) — E-Po, Guitart Manday and
Juarez Fernandez (1966)-E-Y, Ukawa et al. (1966)-E-Pr, Mito et al. (1967)-Pr-J
Presley (1970)-F-Po, Smith (1971)-Po, Aboussouan (1972b)-Pr-Po, Fourmanoir
(1976)-L, Chenet al. (1977)-E, Kendall (1977. l979)-Po, Hussain and Higuchi
(I980)-Y-J
Johnson and Ashe (1984)- Po-J, Leis and Rennis (1983)-Pr-Po
Kendall and Fahay (1979)-Po, Johnson and Ashe (l984)-Po
Kendall (1977, 1979)-Pr-J, Johnson and Ashe (1984)- Po-T
Kendall (1977, l979)-Po, Johnson and Ashe (1 984)- Po-T
HubbsandChu (1934)-T
Kotthaus (1970)-L, Fourmanoir (197 la)- Po, Fourmanoir (1976)-Po, Kendall
(1977. 1979)-Pr-Po
KENDALL: SERRANIDAE
501
Table 123. Continued.
Subfamily
Genus
subgenus
Distnbution
inbc
A-A 1-P
Pikea
+
Rainfordia
+
Grammistini
Aporops
+
Grammistes
+
Grammistops
+
Pogonoperca
+
Pseudogramma
+ +
Rypticus
+ +
Suttonia
+
ELH descnptions
Fourmanoir (1976)— Po
Fourmanoir (1976)— T
Kendall (1977, I979)-Pr-Po. U-is and Rennis (1983)-Pr-Po
Aboussouan (1972b)-Po, Kendall (1977, 1979)-Po
' Randall (1980) includes in Plectranthias: Sayanura, hobuna. Xenanlhias. Pleranthias, Zatanthias. Serranops. Peionlnis. and Zacallanlhias
*• Subdivisions follow Johnson (1983).
' Tortonese (1973) states thai Bertolini (1933b) and Sparta (1935) described Mycteroperca ruba larvae as Ephmephetus alexandnnus and that this mistake has been continued in more recent
literature
rygiophore arrangement (Johnson, 1983). The first and third
dorsal spines and the anal spines are also stout and may bear
serrations. The spine at the angle of the preopercular is elongate
and serrate; there are two smaller spmes dorsal and ventral to
the one at the angle, and these may also bear serrations. There
is a serrate spine on the supracleithrum. The body is "kite-
shaped"; pigment lines the body cavity and there is a large,
conspicuous spot on the caudal peduncle that migrates from the
ventral midline to a midlateral position during flexion.
Grammistini-Liopropomini. — The body is roughly tubular with
a deep caudal peduncle. Among the bones in the opercular series
the preopercular is armed with about five elongated, simple
spines. One or two dorsal fin spines become quite elongate, and
are thin and flexible with pigmented membranous sheaths around
them. Bodies of the larvae are practically devoid of pigment
throughout development.
The following is a summary of the current status of the sys-
tematics and knowledge of larval morphology of each of the
subfamilies of serranids (Table 123).
Serraninae
There has been no revision of this primarily Atlanto-Amer-
ican subfamily, and little work on relationships among species
in the various genera (Bortone, 1977). These are considered the
least specialized of the serranids and are riiainly united by shared
possession of basal percoid characters rather than unique spe-
cializations, which would allow a definitive statement about
monophyly. They possess the four serranid specializations as
mentioned by Johnson (1983), are hermaphroditic or second-
arily gonochoristic (see Kendall, 1977). have a common pre-
dorsal bone pattern (0/0/0/2), and a fairly coherent larval mor-
phology (Fig. 266).
The larvae of Schulcea. Dules. Acanihistius, and Crulmus
are unknown. The following summary of what is known of the
larval morphology of the rest of the serranines is based primarily
on Kendall (1977, 1979). The only more recent contributions
to serranine larval knowledge are the descriptions of Paralabra.x
(Butler et al., 1982).
Centropnstis.— Only one larval type is known, although four
species are named. The eggs and yolk-sac larvae have been
described from reared specimens. Development is typical of
serranines with small simple spines on the preopercular. The
first and second dorsal fins develop at about the same rate; there
are no elongate or armed fin spines. Most pigment is in blotches
in characteristic serranine positions. The body gradually as-
sumes the adult shape.
Paralabra.x— Biil\eT et al. ( 1 982) reared from eggs and described
development of the three species found off California. These
larvae vary from the general serranine pattern of development,
primarily in having pigmented membranes of the pectoral, pel-
vic, first dorsal, and anal fins variously developed among the
species. Pigment is also variously present on the body ventral
to the first dorsal fin.
Serraniculus. — Larvae of the only species (S. pumilio) are deeper
bodied and have more lateral pigment than other serranines.
The flank pigment is composed of three series of dashes (one
along the midlateral septum and one along the base of the dorsal
and anal fins) and superficial small spots over much of the trunk.
The ventral midline spots are small and rather uniform in size.
The first dorsal fin develops concurrently with the second dorsal,
and the spines are no longer than the rays.
Diplectrum— Two distinct types of larvae with the meristic
characters of this genus were found in both Atlantic and Pacific
material. One type (Type 1 ) closely follows the serranine pattern
of development, the main difference being in the early devel-
opment of the spinous dorsal and pelvic fins. The ventral mid-
line pigment spots seem more uniform in size than in other
genera, and there is pigment on some of the fin membranes.
Larvae of the other type (Type 2) are quite different from other
serranine larvae in that the pectoral and pelvic fins develop
early and are enlarged and pigmented on their distal thirds. The
body is practically devoid of pigment except for two spots on
the caudal peduncle— one dorsal and one ventral, and an inter-
nal diffuse area of pigment lateral to the anterior part of the
anal fin that develops after the fin rays are formed.
Serranus. — Larvae of this genus from both sides of the Atlantic
have been described, and reared eggs and yolk-sac larvae were
among the first serranids descnbed (Raffaele, 1888). These lar-
vae differ from the serranid pattern of development in having
eariy-forming elongate dorsal spines and a deeper body. In 5.
cabrilla the pelvic spines develop before any other fin rays and
they and the third dorsal spine become quite elongate. Some of
the smaller ventral melanophores seen in other serranines are
absent from Serranus larvae, while some of the larger spots are
502
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
B
KENDALL: SERRANIDAE
503
more intense. Pigment develops variously at the base of the
dorsal fin and in the membranes of the first dorsal, pelvic, and
anal fins. S. cabrilla has large opposing spots on the caudal
peduncle.
Hypoplectrus— Reared larvae of this genus are quite different
from other serranines. The first dorsal and pelvic fins develop
early and are heavily pigmented. The head and fin membranes
are fleshier than in other serranines, and these larvae do not
possess the characteristic ventral pigment pattern. Rather, there
are a few spots ventral to the base of the first dorsal fin, and a
few blotches ventrally at the base of the pelvic fins, at the anus,
along the base of the anal fin, on the caudal peduncle, and at
the base of the middle of the caudal fin.
Anthiinae
This is a cohesive group of fishes that share several special-
izations in addition to those they hold in common with other
serranids. These specializations include large scales, a highly
arched lateral line, deep bodies and large heads, mainly 10 +
1 6 vertebrae, and a predorsal pattern of 0/00/2 or 0/0/2. They
are generally small, brightly colored reef fishes. The generic
alignments of many species are dubious, and a revision of the
group is badly needed. Most recent work, however, has focused
on describing new species, faunal studies, and some generic
revisions.
Recent and ongoing work (Fitch, 1982; Baldwin, pers. comm.)
has brought out several incongruencies in generic assignments
of Kendall (1977, 1979). In the following summary of what is
known of anthiine larval morphology, generic larval types will
be described, with the understanding that some of the variation
within these may be due to species that are assigned to the genus
incorrectly. Alternate generic placements of species will be noted
as appropriate (Table 1 24). Better definitions of the genera must
await a worldwide revision that will include information on
early life history stages. Larvae of 10 of the 19 currently rec-
ognized anthiine genera are known to some extent (Fig. 267 and
Table 123).
Plectranthias (Fig. 267a). — Randall (1980) included eight nom-
inal genera in this genus, but the monophyly of the included
species is not resolved (W. D. Anderson, Jr., pers. comm., Jan.
1983). Kendall (1977, 1979) described larvae of the American
species (P. garupellus) as having an elongate third dorsal spine,
opposing caudal peduncle pigment blotches as well as a blotch
below the center of the first dorsal fin, and no serrated head or
fin armature (rather the characteristic anthiine spines are thin
and weakly developed). The larvae showed the least develop-
ment of anthiine larval characters among American genera.
Anthias (¥\%. 267c). — (includes Pronotogrammus multifasciatus
(see Fitch, 1982)) This is a speciose circumtropical genus that
has provisionally been subdivided into three subgenera (Randall
and Lubbock, 198 1). Larvae of several species from around the
world have been described. They share a number of larval char-
Table 124. ReassionmentofSome Anthiine Larvae. Those of Ken-
dall (1977, 1979) reassigned by Baldwin (pers. comm.) and Kendall,
based on work on adults from the eastern Pacific by Fitch (1982) and
from the western Atlantic by Anderson and Heemstra (1980) and W.
D. Anderson (pers. comm.. unpublished data). Letters after most likely
species names refer to Baldwin (B) and Kendall (K) who recognized
these reassignments.
Kendall. 1977, 1979
Most likely species
Figure
Pronotogrammus
aureoruhens
Pronologrammus eos
Anihias gordensis
Amhias sp. Type 2
Hemanthias peruanus
Hemanthias leptus—B 267f
Hemanthias signifer— B
Pronotgrammiis multi-
fascialus—K.
Holanthias martinicensis—B 267d
Pronotogrammus eos— K
acters, but there are some notable differences among the species.
The second or third dorsal spine is elongate and thin (the first
may be late forming, so the elongate spine may always be the
third); the first few dorsal spines and the pelvic spine are early
forming; the elongate dorsal spine has a pigmented sheath; the
preopercular and interopercular have long serrate spines; and
there are generally two pigment spots ventrally on the caudal
peduncle. There is a simple supraoccipital spine in some species
and a variable number of spines on a ridge above the eye.
Pigment, in addition to that mentioned above, varies among
the species and some species become fully scaled during the
larval stage. Whether these diflTerences in larval characters can
be related to the subgeneric alignment of species must await
further larval descriptions. Fitch (1982) synonymized the Pacific
Anthias (A. gordensis), whose larvae Kendall (1977, 1979) de-
scribed, with Pronotogrammus multifasciatus.
Franzia. — Eggs and yolk-sac larvae of F. squamipmnis have
been described (Suzuki et al., 1978) but later larval stages are
unknown.
Caesioperca. — Vo\xrmano\r(\916) illustrated the head and brief-
ly described a transforming specimen thought to belong to this
genus. It has a smooth supraoccipital region and no spiny ridge
above the eye, but has simple stout spines in the characteristic
position on the preopercular and interopercular. The informa-
tion presented is too brief for further evaluation of anthiine
larval characters.
Luzonichthys. — Fourmanoir (1976) illustrated the anterior por-
tion and briefly described two transforming specimens of this
genus. These have probably lost some of their larval characters,
since the mouth is already subterminal and the body covered
with scales. The spines on the preopercular are not especially
elongate, but one on the interopercular is pronounced, simple,
and stout. Anterior dorsal fin spines appear thin and not pro-
duced.
Fig. 266. Examples of serranine larvae: (A) Centropristis striata, 8.3 mm, from Kendall ( 1 979); (B) Paralabrax clathratus. 7.4 mm, from Butler
et al. (1982); (C) Serranicutus pumilio. 5.8 mm, from Kendall (1979); (D) Diplectrum sp., 6.1 mm, from Kendall (1979); and (E) Serranus sp.,
5.5 mm, from Kendall (1979).
Fig. 267. Examples of anthiine larvae. (A) Plearanthias garupelhis. 5.5 mm, from Kendall (1979); (B) Pronotogramumus aureoruhens, 9.8
mm, original illustration; (C) Anihias sp., 5.3 mm. from Kendall (1979); (D) Holanthias manmicensis. 8.4 mm, from Kendall (1979), labelled
Anlhias sp. Type 2; (E) Hemanthias vivanus. 6.8 mm, from Kendall (1979); and (F) Hcmanthias leptus. 6.0 mm, from Kendall (1979) labelled
Pronotogrammus aureoruhens.
506
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
KENDALL: SERRANIDAE
507
5acMra. — Reared eggs and yolk-sac larvae were described by
Suzuki et al. ( 1 974) and a postflexion larva, illustrated and brief-
ly described by Fourmanoir(1976). shows characters of anthiine
larval development. The latter specimen has the third dorsal
and pelvic spines extremely elongate and with a pigmented
sheath; the opercular and interopercular are armed with stout
serrate spines, and there is a similar more ventrally-directed
spine anterior to these; the anal spines are stout and serrate;
there is a serrate ridge above the eye, and a midlateral pigment
dash on the caudal peduncle.
Pronotogrammus-Hemanthias.— As presently understood, two
species assigned to each of these genera occur in the eastern
Pacific (P. COS. P. multifasciatus, H. signifer, and H. peruanus,
see Fitch ( 1 982)), and there are two Hemanthias and one Prono-
togrammus in the western Atlantic (H. leptus. H. vivanus, and
P. aiircoruhens).
Kendall (1977, 1979) assigned larval types from both oceans
to these genera. More recently, Baldwin (pers. comm.) has es-
tablished alternate generic assignments for some of Kendall's
(1977, 1979) types based on more complete meristic data and
has assigned a previously undescribed type to Pronotogrammus
aureorubens. Thus, present generic assignments, do not coincide
with the larval types described by Kendall (1977, 1979). In the
following, the morphology of the larval types of Kendall (1977,
1979) will be summarized under the species whose larvae are
represented by these types.
Hemanthias signifer, Hemanthias leptus (Kendall's Pronoto-
grammus eos and P. aureorubens) (see Fig. 2670. — These larvae
are characterized by serrate, spiny armature in the opercular
region, supraoccipital crest simple or absent, first spines of the
dorsal fin and the pelvic fin early developing but not becoming
elongate or serrate, and midlateral trunk pigment.
Pronotogrammus eos. Hemanthias vivanus (Kendall's Heman-
thias peruanus and H. vivanus) (see Fig. 267e).— These larvae
develop a complex "cockscomb" ridge on the supraoccipital, a
serrate ridge above the eye, some serrate spines on the pre-
opercular and interopercular, some serrate fin spines (in all spiny
rayed fins in H. vivanus, only in the pelvic of P. eos), and spiny
scales.
Pronotogrammus aureorubens (Fig. 267b). — Baldwin (pers.
comm.) has found larvae from the western Atlantic that are
heavily spined and possess the meristic characters of P. aureo-
rubens. These larvae are completely scaled, have serrations on
spines of all spinous fins which are also quite stout, and have
heavy serrate spines in the opercular region. The dorsal aspect
of the head is covered with spinous ridges including a complex
cockscomb spine on the supraoccipital. There are four blotches
of pigment dorsally on the body: two ventral to the first dorsal
fin, one ventral to the second dorsal fin, and one on the caudal
peduncle.
Holanthias (Fig. 267rfA-Kendall (1977, 1979) illustrated and
briefly mentioned an anthiine larva he called .4nthias sp. Type
2 which has been shown to be Holanthias martinicensis (Bald-
win, pers. comm.). These larvae are deep-bodied with large
heads and mouths. They develop serrate spines in the opercular
region, and a simple supraoccipital spine in post-flexion larvae.
They have several spines above the eye and develop scales dur-
ing the larval stage. They have some pigment in the membrane
of the first dorsal fin as well as a line on the body ventral to the
second dorsal fin. Baldwin (pers. comm.) has pointed out the
similarities between Holanthias martinicensis larvae and those
Kendall (1977, 1979) described as .4nthias gordensis, including
the early appearance of scales, not noted by Kendall (1977,
1979).
Selenanlhias.—A transforming specimen illustrated and briefly
described by Fourmanoir (1973) is deep-bodied but has no elon-
gate fin spines. It appears to be fully scaled and has stout, pos-
sibly serrate preopercular and interopercular spines.
Epinephelinae
Johnson (1983) has dealt with the systematics of several gen-
era that had been thought variously related to each other. These
are mainly genera in the epinepheline-grammistine lineage of
Kendall (1976). On the basis of several characters, Johnson
proposed that these genera form a monophyletic lineage
(subfamily Epinephelinae) that is composed of five tribes (Ni-
phonini, Epinephelini, Diploprionini, Liopropomini, and
Grammistini). Some early life history stages are known for all
of the tribes except Niphonini (Fig. 268). The larvae share the
elongation of one or two anterior dorsal spines, and the larvae
and adults share predorsal bone and pterygiophore arrange-
ments which presumably function to support the larval dorsal
spines (Johnson, 1983). In the Epinephelini, the dorsal spines
are stout and serrate, whereas in the other three tribes they are
extremely elongate, flexible, and some have siphonophore-mim-
icking pigment and shape.
The following is a summary of what is known of the mor-
phology of early life history stages of fishes in the epinepheline
tnbes of Johnson (1983).
J^iphonini. — Niphon spinosus. the sole member of this tribe, has
unknown larvae but Johnson (1983) speculated that on the basis
of first dorsal pterygiophore morphology and presumed rela-
tionships, their third dorsal spine should be elongate.
Epinephelini.— Larvae are known only for those genera occur-
ring in Atlanto-American waters. Several species have been
reared and their egg and larval development described (see Table
123).
Epinephelus. — Larvae of species from every ocean belonging to
this circumtropical genus are known. Smith (1971) placed the
American members of the genus in five subgenera: Epinephelus,
Promicrops, Cephalopholis, Dermatolepis. and .-ilphestes. These
had formerly been considered genera, and members of these
occur in other parts of the world. Johnson and Ashe ( 1 984) were
able to identify larvae of most species of American Epinephelus
primarily on the basis of spinelets on the elongate dorsal and
pelvic spines. They compared spinelet patterns among members
of the subgenera and species groups of Smith (1971) and found
Fig. 268. Examples of epinepheline larvae: (A) Epinephelini: Paranthias furcifer. 8.6 mm. from Kendall ( 1 979); (B) Liopropomini: Liopropoma
sp., 1 1.0 mm. Collected by G. R. Harbison, 16 May 1981, b^Sl.S'S, 150°21.8'E; and (C) Grammistmi: Ryplicus sp., 6.6 mm, from Kendall (1979).
508
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
that most share common patterns (e.g., species groups E. sthatus
and E. adscensionis), although there are some notable problems
(subgenera Cephalopholis and Alphestes). Thus, in Epinephehis
there is general concordance between the only distinguishing
characters of the larvae (spinelet patterns) and the relationships
hypothesized based on a variety of adult characters; but thor-
ough analysis must be done to resolve apparent discrepancies.
Paranthias. — One species (P.furcifer) occurs in American waters
of the Atlantic and Pacific Oceans. The larvae have a unique
spinelet pattern on the dorsal fin spines, and have internal no-
tochord pigment not found in other epinephelines (Johnson and
Ashe, 1984). This genus as an adult is quite distinct ecologically
and morphologically.
Mycteroperca.— This American genus with 13 species is distin-
guished from the other epinepheline genera by several charac-
ters, including usually having more anal rays (11-13). The species
of Mycteroperca cannot be distinguished as larvae, and their
spinelet patterns resemble those of several members of Epineph-
elus (e.g., E. niveatus, E. flavolimbatus, and E. acanthistius).
However, Mycteroperca larvae have a melanophore at the
cleithral symphysis, which is not found in any of these species
oi Epinephelus (Johnson and Ashe, 1984).
Gon/op/ec/rus. — Postflexion larvae of the only species, Gonio-
plectrus hispamis, are known (Kendall and Fahay, 1979). The
larvae are more robust and have shorter elongate dorsal and
pelvic spines than other American epinephelines. Also, these
elongate spines are different in cross section and spinelet ap-
pearance than those of other epinephelines (Johnson and Keen-
er, 1984).
Diploprionini. — A photograph of a transforming larva, a draw-
ing of a juvenile, and a brief description of the juvenile showed
fish with long flexible dorsal spines and rather deep bodies (Hubbs
and Chu, 1934). The second and third dorsal spines are ex-
tremely produced in the larva, but only the third is in the ju-
venile. The photograph of the larva does not allow more detailed
observation.
Liopropomini. — Larvae of Liopropoma/ Pikea are known and
cannot presently be distinguished on the basis of larval char-
acters (Kendall, 1977, 1979). They were first described as a new
genus, Flagelloserranus. by Kotthaus ( 1 970). Jeboehlkia is known
from a single, small specimen which shows traits of being a
transforming larva (Robins, 1967).
Lioproma/Pikea. — The general body shape is similar to that of
the serranines, although the gut is shorter and there is a space
between the anus and the origin of the anal fin. The caudal
peduncle is both longer and deeper than it is in serranines. The
most outstanding developmental feature is the presence, even
in small larvae, of two elongate, thin dorsal spines. These de-
velop before other fin rays, reach a length of up to three times
the fish length, and become the second and third dorsal spines.
These spines are delicate and are broken in many specimens.
Kotthaus ( 1 970) described the presence of thick tissue surround-
ing these spines; the tissue around the second spine has two
vane-like swellings on its distal third and the tissue around the
third spine is tubular for its entire length. The distal portion of
both spines is pigmented with several large melanophores. The
remaining fin rays develop their adult proportions without any
pronounced elongations. The ventral fins develop more slowly
than those of most other serranids.
Except for the pigment on the elongate dorsal fin spines, most
larvae are unpigmented. Some spots develop on the hindbrain
surface in larger larvae, probably representing the onset of ju-
venile pigment.
Jeboehlkia.— The single species {J. gladifer) is known only from
the holotype, a 40.8 mm female. Characters that indicate that
it may not have completed transformation, or may be paedo-
morphic, include the virtual lack of pigment, the enlarged eye,
and the elongate first dorsal spine (see Robins, 1967).
Grammistini. — Fishes in this tribe have been variously grouped
as members of families separate from the serranids and as
subfamilies of the serranids. Larvae of four of the seven genera
placed in this tribe by Johnson (1983) are known. The first or
second dorsal spine is elongate and flexible, and the preopercular
margin is armed with about five subequal spines in larvae of all
four genera.
Grammistes.—A single, 1 1 mm postflexion larva of G. se.xiline-
atus illustrated by Fourmanoir (1976) has an elongate flexible
first dorsal spine and five spines on the preopercular margin. It
is well developed, rather deep-bodied, and appears to lack pig-
ment except on the pectoral fin which is covered with fine me-
lanophores on its distal third.
Aporops.— The anterior portion of a 12 mm postflexion larva
of Aporops bilinearis illustrated by Fourmanoir (1976) has the
first dorsal spine elongate and flexible and five spines on the
preopercular margin. It is well developed and is not as deep-
bodied as the aforementioned Grammistes larva. No pigment
is evident in the illustration.
Pseudogramma. — A developmental series of P. gregoryi was
described by Kendall (1977, 1979) and Leis and Rennis (1983)
illustrated a series of P. polyacantha. These larvae have shallow
tubular bodies; a greatly elongate, flexible dorsal spine (the first
or second); precocious enlarged pectoral fins; a gap between the
anus and the anal fin; and a general lack of pigment except on
the pectoral fin of small larvae and on the sheath that surrounds
the elongate dorsal spine.
i?.V77;;cM.s. — Aboussouan (1972b) illustrated and briefly de-
scribed two larvae, and Kendall (1977, 1979) compared these
with specimens he described from the western Atlantic. These
larvae have the first dorsal spine produced, flexible, and sur-
rounded by a pigmented sheath; about five preopercular spines;
an enlarged pectoral fin that may be pigmented; rather long rays
in the second dorsal, caudal, and anal fins; small, late-developing
pelvic fins; a lack of body pigment; and are moderately deep-
bodied at the nape.
Relationships
Although known larvae of serranids show a diversity of char-
acters that will probably permit them to be used in definitive
studies of relationships within the group, such studies are pres-
ently premature (Fig. 269). More characters need to be traced
ontogenetically, and larvae of more species, particularly in the
Anthiinae and several tribes of Epinephelinae, need to be de-
KENDALL: SERRANIDAE
509
preopercular
interopercular
preopercu
interopercular
B
preopercu
nteropercular
D
preopercular
interopercular
Fig. 269. Representative preopercular and interopercular bones from larval serranids (from Kendall, 1979); (A) Serraninae: Serranus sp.; (B)
Anthiinae: Anthias sp. Type 1; (C) Epinephelinae: Epinephelini, Epmephelus nivealus: and (D) Epinephelinae: Grammistini. Pseudogramma
gregoryi.
scribed. At present, however, some statements can be made
concerning serranid systematics from what is known about the
larvae.
The serranid subfamilies are clearly distinct as larvae. In fact,
it is not possible to characterize the Serranidae based on larval
morphology, because no characters unite the subfamilies while
separating them from larvae of all other families. Serraninae
larvae seem to be the least specialized and are more similar to
percoid genera thought to represent the basal stock from which
serranids arose (e.g., Morone. Lateolahrax. and Dicentrarchus).
The serranine genera can be distinguished from each other and
ordered in a rough progression of divergence from the supposed
ancestral larval form (as exemplified by Morone), as follows:
Serraniculus. Centroprislis-Paralabrax, Diplectrum Type 1,
Serranus (see Kendall, 1979). Characters that lead to this as-
sessment include pigment, body shape, sequence of dorsal spine-
soft ray development, and dorsal fin spine elongation.
Based on larval and other evidence, it appears that two major
radiations from the ancestral serranines arose leading to the
anthiine and the epinepheline lineages. The anthiines form a
fairly cohesive group of fish which are at the same time quite
speciose. The generic alignment of many anthiines is unclear
and in some cases larval evidence is in conflict with that based
on adults. Anthiine larvae, like the adults, share several char-
acters that unite them, yet they are quite diverse and will prob-
ably prove to be excellent subjects for phylogenetic investiga-
tions. Larvae of only about half of the presently understood
anthiine genera are known to any extent, some of them only
from one transforming larva. Thus the lack of generic revisions
and incomplete knowledge of larval development makes it pres-
ently unreasonable to attempt a thorough systematic assessment
that would include larvae. Within the group, a progression of
increasing spinyness and armature is apparent. Among the lar-
vae described to date, armature seems to be added as follows:
510
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
elongate preopercular and interopercular spines, serrate pre-
opercular and interopercular spines, stout pelvic and first three
dorsal spines, supraoccipital spine, serrate dorsal and pelvic
spines, serrate head spines on several bones, and spiny scales
developing during the larval stage.
The other major line of divergence from the serranines is the
five tribes of the epinephelines. Johnson (1983) pointed out the
adult features that characterize this subfamily and the tribes
within it, although he did not provide a detailed analysis of the
relationships among the tribes. The larvae (representatives of
four tribes are known) all have one or two quite elongate dorsal
spines. In the Epinephelini, the elongate dorsal spines are stout
and serrate; in the other tribes, they are flexible, thin, and in an
elaborately pigmented sheath. Thus it appears from the larvae
that the Diplopionini, Liopropomini, and Grammistini may
form a monophyletic group within the Epinephelinae.
Epinephelini larvae are all quite similar but some genera can
be separated by larval characters (Gonioplectrus, and Paran-
ihias), although larvae are unknown for several genera. Gon-
ioplectrus larvae are most similar to anthiine larvae and may
represent the most primitive extant epinephelini state. Johnson
(1983) suggested that Niphon represented the primitive sister
group of all other epinephelines and that its unknown larvae
may have an elongate third rather than second dorsal spine.
There is less variation in size of the second and third dorsal
spines in Gonioplectrus, compared to other Epinephelini, which
adds credence to the above suggested relationships.
Few larval representatives of the other epinepheline tribes
[grammistine lineage of Kendall (1976)] are known and none
of them have been studied in detail. Their elongate, pigmented
flexible dorsal spines, lack of corresponding elongate pelvic
spines, five subequal preopercular spines, and dearth of body
pigment unite the known larvae. Larvae of Diploprion are rather
deep-bodied compared to the more tubular bodies of the other
known larvae grammistines. The second and third dorsal spines
are produced in Diploprion and Liopropoma. but only one spine
is produced in members of the Grammistini. In this group of
serranids there appear to be larval characters that will be helpful
in systematic studies, but larvae of more representatives must
be known in more detail before such studies will be meaningful.
National Marine Fisheries Service, Northwest and Alaska
Fisheries Center, 2725 Montlake Boulevard East,
Seattle, Washington 981 12.
Carangidae: Development
W. A. Laroche, W. F. Smith-Vaniz and S. L. Richardson
THE family Carangidae (jacks, trevallys, and pompano) has
traditionally been assigned to the suborder Percoidei, an
assemblage of generalized perciform fishes (Lauder and Liem,
1983). The family is notably heterogenous, including species
which differ widely in structure and appearance. Phylogenetic
relationships within the suborder and even the familial limits
of the Carangidae are not clearly established (see Smith-Vaniz,
this volume). The family is composed of approximately 140
species and 30 genera (Table 1 25) many of which remain poorly
defined.
Carangids are found world-wide in tropical and warm tem-
perate marine and estuarine waters. Carangids are actively
swimming fishes which range from small schooling planktivores
to large solitary piscivores (Berry and Smith-Vaniz, 1978). Some
species of carangids are known to spawn pelagically offshore,
i.e., Seriola lalandi = S. dorsalis (Baxter, 1960) and Trachurus
symmetricus (Ahlstrom and Ball, 1954), while others spawn
close to shore and near the bottom, i.e., Caranx ignobilis (von
Westemhagen, 1974) and Oligoplites saurus (Aprieto, 1974).
The greatest amount of information concerning early life stages
exists for species of Decaplerus and Trachurus on which research
has focused due to their commercial importance.
Development
Eggs
Carangids have spherical, pelagic eggs which have a narrow
perivitelline space and range in diameter from about 0.7 to 1.3
mm. One to several oil globules are usually present, and egg
envelopes are clear, unsculptured, and lack filaments (Ahlstrom
and Ball, 1954; Miller and Sumida, 1974; James, 1976a). The
eggs of Naucrates ductor have erroneously been reported to be
demersal, adhesive, with a fine entangling filament at one pole
(Gilchrist, 1918) and attached to sharks and the hulls of ships
(Gilchrist, 1918; Shuleikin, 1958). They are actually pelagic,
non-adhesive, and without filaments (Barnard, 1926; Sanzo,
1931a; Maksimov, 1969).
Development proceeds in the typical manner of pelagic fish
eggs (Ahlstrom and Ball, 1954; Miller and Sumida, 1974). Eggs
hatch 24 to 48 hours after spawning at water temperatures be-
tween 18 and 30 C° (temperature range within which eggs and
larvae are most commonly taken).
Carangid eggs are similar in size and appearance to those of
many other marine fishes. Thus, identification even to family
level may be difficult or frequently impossible using presently
known characters.
Larvae
A/or/)/;o/(7gi'. — Information is available on at least one devel-
opmental stage for 58 of the 140 valid species representing 24
of 30 genera (Table 125). However, even among those taxa for
which descriptive information is available, inconsistent quality
in descriptive text and coverage of the developmental period
make detailed morphological comparisons and identifications
based upon these descriptions difficult in many cases. Laroche
et al. (MS) have refined developmental terminology for caran-
gids so as to define developmental stages more precisely and
thus improve comparability of descriptions between taxa.
LAROCHE ET AL.: CARANGIDAE
511
Table 125. Species and World Distribution List for the Family Carangidae. Selected literature references deal with descriptions of larvae
and juveniles.
Species
Ind.'
Wesl
Cent*
East
Wesl
Easl
Ocean
Pac.
Pac
Pac,
All.
All.
References*
Aleclis alexandnniis (E. Geoffrey St.-Hilaire)
Alectis cilians (Bloch)
Aleclis indicus (Riippell)
Alepes djedaba (ForsskSl)
Alepes melanoptera Swainson
Alepes sp.
Alepes vari (Cuvier)
Atropus alropos (Bloch and Schneider)
Atule male (Cuvier)
Campogramma glaycos (Lacepede)
Carangoides armaliis (Riippell)
Carangoides bajad (ForsskSl)
Carangoides bartholomaei (Cuvier)
Carangoides caeruleopinnalus (Riippell)
Carangoides chrysophrys (Cuvier)
Carangoides dinema Sleeker
Carangoides equula (Temminck and Schlegel)
Carangoides ferdau (ForsskSl)
Carangoides fulvoguttatus (ForsskSl)
Carangoides gymnosielhus (Cuvier)
Carangoides hedlandensis (Whitley)
Carangoides humerosus (McCulloch)
Carangoides malabaricus (Bloch and Schneider)
Carangoides oblongus (Cuvier)
Carangoides orthogrammus Jordan and Gilbert
Carangoides olrymler (Jordan and Gilbert)
Carangoides plagiolaenia Sleeker
Carangoides praeustus (Sennett)
Carangoides ruber (Bloch)
Carangoides talamparoides Sleeker
Carangoides uii (Wakiya)
Carangoides vinclus (Jordan and Gilbert)
Caranx bucculentus Alleyne and Macleay
Caranx caballus Giinther
Caranx caniniis Giinther
Caranx crysos (Mitchill)
Caranx hippos (Linnaeus)
Caranx ignobilis (ForsskSI)
Caranx lalus Agassiz
Caranx lugubris Poey
Caranx melampygus Cuvier
Caranx papuensis Alleyne and Macleay
Caranx sem Cuvier
Caranx senegallus Cuvier
Caranx sexfasciatus Quoy and Gaimard
Caranx tille Cuvier
"Caranx" kohcru Hector
"Caranx" para Cuvier
"Caranx" rhonchus E. Geoffroy St.-Hilaire
Chloroscomhrus chrysurus (Linnaeus)
Chloroscombrus orqueta Jordan and Gilbert
Decaplcrus kurroides Sleeker
Decapteriis macarellus (Cuvier)
Decaplcrus macrosoma Blecker
Decaplcrus maruadsi (Temminck and Schlegel)
Decaplcrus muroadsi (Temminck and Schlegel)
Decaplcrus punctalus (Cuvier)
+
+
-
+
+
-
+
+
-
-
+
-
+
+
-
+
+
-
+
+
+
+
+
-
+
+
:
+
+
_
+
+
-
+
+
-
+
+
+
+
+
+
+
+
-
+
+
-
+
+
-
-
+
-
+
+
-
+
+
-
+
+
+
+
+
+
+
+
-
+
+
—
+
+
—
+ *Aboussouan. 1975
+ 'Aboussouan, 1968a: *Fowler, 1936;
Ginsburg 1952; *Johnson, 1978
•Tsokur, 1977
*Kuthalingam, 1959a; *Miller and Sumida,
1974; •Miller et al.. 1979; Zvyagina and
Rass, 1977
•Berry, 1959b; •Johnson 1978; Uroche et
al. (in prep.)
+
+
+
+
+
+
+
-
+
+
+
-
+
-
+
-
+
-
—
+
+
+
+
+
+
+
-
-
+
+
—
•Berry, 1959b
+ Aboussouan, 1975; Berry, 1959b; Johnson,
1978; Montolio. 1976; McKenney et al.,
1958
+ Berry, 1959b; Johnson, 1978
Berry, 1959b; Johnson, 1978
+ Aboussouan, 1975
- Ahlstrom and Sumida (in prep.)
•Bapat and Prasad, 1952
Aboussouan, 1967; Aboussouan, 1975; Co-
nand and Franqueville 1973
Aboussouan, 1968a; Aboussouan, 1975;
Laroche et al. (in prep.)
Ahlstrom and Sumida (in prep.)
?Delsman, 1926a
Shojima, 1962
■Aboussouan, 1975; Aprieto. 1974; Hilde-
brand and Cable, 1930; Johnson, 1978;
Montolio, 1976
512
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 125. Continued.
Genus'
Species
Ind.'
Ocean
Wcsl
Pac.
Cent.'
Pac
East
Pac
West
All,
East
Atl.
Decapterus russelti (Riippell)
Decaptenis scomhrinus (Valenciennes)
Decapterus labl Berry
Decapterus n. sp. "stonebrass scad'"
Elagatis bipinnulata (Quoy and Gaimard)
Gnathanodon speciosus (ForsskSl)
Hemicaranx amhiyrhynchus (Cuvier)
Hemicaranx bicolor (Gunther)
Hemicaranx leucurus (Gunther)
Hemicaranx zelotes Gilbert
Lichia amia (Linnaeus)
Magalespis cordyta (Linnaeus)
Naucrates ductor (Linnaeus)
Otigoplites altus (Gunther)
Oligoplhes patometa (Cuvier)
Otigoplites refidgens Gilbert and Starks
Otigoplites saliens (Bloch)
Otigoplites saurus (Schneider)
Panlotahus- radiatus (Macleay)
Parastromateus niger (Bloch)
Parana signata (Jenyns)
Pseudocaranx clulensis (Guichenot)
Pseudocaranx dentex (Bloch and Schneider)
Pseudocaranx nrigfiti (Whitley)
Scomberoides commersonianus Lacepede
Scomberoides lysan (ForsskSl)
Scomberoides tala (Cuvier)
Scomberoides tot (Cuvier)
Selar boops (Cuvier)
Setar crumenophlhatmus (Bloch)
Selaroides leptolepis (Cuvier)
Selene brevoortii (Gill)
Selene brownii (Agassiz)
Selene dorsalis (Gill)
Selene oerstedii Lutken
Selene peruviana (Guichenot)
Selene setapinnis (Mitchill)
+
+
+
Selene vomer (Linnaeus)
Seriola carpenteri Mather
Seriota dumerili (Risso)
Serwta fasciata (Bloch)
Seriola hippos Gunther
Seriola latandi Valenciennes
Seriola peruana Steindachner
Seriola quinqueradiata Temminck and Schlegel
Seriola rivoliana Cuvier
+
+
+
-
—
+
+
-
+
+
+
+
+
+
+
+
+
+
+
_
_
—
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ +
- +
+
?*Delsnian, 1926a; ?*Tsokur, 1977; Vijay-
araghavan, 1958
Aprieto, 1974; Berry, 1969; Johnson, 1978;
Laroche et al. (in prep); Okiyama, 1970
Ahlstrom and Sumida (in prep.); Miller et
al., 1979
Hoese and Moore, 1977; Laroche et al. (in
prep.)
Lo Bianco, 1909; Padoa, 1956c
?Kuthalingam, 1959a
Ahlstrom and Sumida (in prep.); Lutken,
1880; Padoa, 1956c; Pertseva-Ostroumova
and Rass, 1973; Roule and Angel, 1930;
Sanzo, 1930a, 1931
Aprieto, 1974; Johnson, 1978; Laroche et
al. (in prep.)
Phonlor, 1979
♦James, 1976a; *Padoa, 1956c; ?*Schnaken-
beck, 1931
•Premalatha, 1977
*Delsman, 1926a; *Miller et al., 1979; Zvy-
agina and Rass, 1977
?Bapat, 1955
Ahlstrom and Sumida (in prep.)
Laroche et al. (in prep.)
*Aboussouan, 1975; *Conand and Franque-
ville, 1973
Ahlstrom and Sumida (in prep.)
*Fowler, 1936; *Ginsburg, 1952; Johnson,
1978; Laroche et al. (in prep.); Lutken,
1880
Aprieto, 1974; Fowler, 1936; Ginsburg,
1952; Johnson, 1978; Laroche et al. (in
prep.); Lutken, 1880
?Hildebrand and Cable, 1930; Johnson,
1978; Laroche et al. (in prep.); Padoa,
1956c; Roule and Angel. 1930; Sanzo.
1930c, 1933b
Ginsberg, 1952; Johnson, 1978; Laroche et
al. (in prep.)
+ Ahlstrom and Sumida (in prep.); ?Brownell,
1979
Lutken, 1880; Mitani, 1960; Uchida, Dotsu
et al., 1958
+ *Ginsburg. 1952; Laroche et al. (in prep.)
LAROCHE ET AL.: CARANGIDAE
513
Table 125. Continued.
Species
Ind '
Ocean
West
Pac.
Cenl'
Pac
East
Pac
West
Atl.
East
Atl.
Seriola zonata (Mitchill)
Seriolina nigrofasciala (Ruppell)
Trachinotus afncanus Smith
Trachinolus anak Ogilby
Trachinotus baiUonii (Lacepede)
Trachinolus blochii (Lacepede)
Trachinotus carolinus (Linnaeus)
Trachinotus cayennensis Cuvier
Trachinotus fatcatus (Linnaeus)
Trachinolus goodei Jordan and Eveimann
Trachinotus goreensis Cuvier
Trachinotus kennedyi Steindachner
Trachinolus marginatus Cuvier
Trachinolus ma.xillosus Cuvier
Trachinotus mookalee Cuvier
Trachinolus ovatus (Linnaeus)
Trachinolus paitensis Cuvier
Trachinotus rhodopus Gill
Trachinolus russelii Cuvier
Trachinotus slilbe (Jordan and MacGregor)
Trachinotus terala Cuvier
Trachinolus velox Ogilby
Trachurus declivis (Jenyns)
Trachurus delagoa Nekrassov
Trachurus japonicus (Temminck and Schlegel)
Trachurus indicus Nekrassov
Trachurus ialhainl Nichols
Trachurus incditcrraneus (Steindachner)
Trachurus murphyl Nichols
Trachurus novaezelandiae Richardson
Trachurus picluratus (Bowdich)
Trachurus syinmelricus (Ayres)
Trachurus trachurus (Linnaeus)
Trachurus trecae Cadenat
Ulua aurochs (Ogilby)
Ulua nwnlalis Cuvier
Uraspis hclvola Forster
Uraspis secunda Poey
Uraspis uraspis Giinther
+
+
+
+
+
+
+
+
+
+
-
-
+
+
+
+
+
+
+
+
+
+
—
—
Aprieto, 1974; Ginsburg, 1952; Johnson,
1978; Lutken, 1880
Fields, 1962; Johnson, 1978; Laroche et al.
(in prep.)
Fields, 1 962; Hildebrand and Schroeder,
1928; Johnson, 1978; Laroche et al. (in
prep.)
•Fields. 1962; Johnson. 1978; Laroche et al.
(in prep.)
Aboussouan, 1975
•De Gaetani, 1940; ♦Padoa, 1956c
Shojima, 1962; Uchida et al., 1958
♦Tsokur, 1977
*de Ciechomski and Weiss, 1973; Johnson,
1978; Laroche et al. (in prep.)
Demir, 1961; Padoa, 1956c; Sanzo, 1932a
Santander and de Castillo, 1971
Aboussouan, 1975
Ahlstrom and Ball, 1954; Ahlstrom and
Sumida (in prep.)
Aboussouan, 1975; Arbault and Boutin,
1968c; Brownell, 1979; Demir, 1961; Eh-
renbaum, 1905-1909; Haigh, 1972b; Kili-
achenkova, 1970; Kingetal., 1977; Leta-
connoux, 1951; Padoa, 1956c; Russell,
1976; Schnakenbeck, 1931
Aboussouan, 1967; Aboussouan, 1975
Johnson, 1978
' Carangid getienc limits arc not well established and some taxa here rccogni/ed ultimately may be allocated to subgenenc status. Carnn^otdes is a poorly defined group that may include
several subunits worthy of recognition- The three species assigned to "Carunx" are not closely related and their generic placement is uncertain.
- Panlntahus Whitley, 1931 is here recognized as a senior synonym oi Absalom Whitley. 1937. Recent examination of the syntypes oi Caranx parasilus Garman {type-species of Panlolabus)
has revealed that they are conspecific with C radiants Macleay. type-species of Absalom
' Species that reach their western distnbutional limit on the eastern margin of the Indian Ocean (including western Australia) are not tabulated as occumng in the Indian Ocean
* Species that reach their eastern distnbutional limit on the western margin of the Pacific Plate (see Springer. 1 982) are not tabulated as occumng in the Central Pacific; Easier Island is treated
as a component of the central Pacific.
' Aslensk indicates scientific name used in cited reference differs from present allocation; question mark indicates only a provisional identification given in cited reference, or adult taxonomy
of group so inadequate at time of publication that specific identification must be treated as suspect
Development in carangids proceeds relatively directly to-
wards the adult stage. Adult characters are gradually acquired
without remarkable, sudden metamorphoses (developmental rale
changes) occumng between stages.
Carangid lai-vae are relatively small and undeveloped at
hatching, usually 1.0 to 2.0 mm notochord length (NL), with a
relatively large yolk sac. Head size, presence of 24-27 myo-
meres, and possession of an oil globule at the anterior of the
514
ONTOGE>fY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 270. (A) Flexion larva (5.4 mm) of Trachiirus lalhaini: postflexion larvae (5.5, 5.6 mm) of (B) Decaplerm punclatus and (C) Selar
crumenophthalmus; and (D) early flexion larva (4.6 mm) of Chloroscombrus chrysurus.
Fig. 271. (A) Early flexion lar\a (3.1 mm) of Aleclis ciliaris; (B) postflexion lana (4.9 mm) of A tide male: and (C) flexion larva (4.0 mm) of
Gnathanodon speciosus.
516
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
B
»T^;
■f'-' ™*'*
,^-J^
^
Fig. 272. (A) Postflexion larva (5.9 mm) of Trachinolus caroUnus; (B) late flexion larva (4.7 mm) of Naucrales ductor. and (C) poslflexion
larva (5.3 mm) of Scnmberoides lysan.
LAROCHE ET AL.: CARANGIDAE
517
Fig. 273. Late postflexion larvae oi (\) Elagatis bipinnulata (1 1.4 mm); (B) Oligoplites saurus (8.6 mm) and (C) Seriola zonata (9.5 mm)
yolk sac, ventral to the head, are the most outstanding characters
of yolk sac larvae. The mouth is not formed, and the gut is
undeveloped. Eyes lack melanistic pigmentation; fins are un-
developed; the notochord is straight; and head spines are lacking
(Ahlstrom and Ball, 1954; Aprieto, 1974; Miller and Sumida,
1974). The present state of knowledge is not adequate to estab-
lish a set of characters which will distinguish pre-fin formation
carangid larvae from larvae of all other marine fish families in
the world. Newly hatched carangid larvae are difficult to identify
even to family due to the paucity of diagnostic morphological
characters and multitude of perciform taxa which co-occur and
have similar-appearing larvae. Since larvae of many taxa remain
unknown, the problem is even more complicated. However,
within restricted and well-defined geographic areas it may be
possible to define such a character set if the fish fauna is well
known (Laroche et al., MS).
Following yolk absorption, larval carangids range from rel-
atively slender forms, i.e., body depth (BD) 20 to 27% SL in
Oligoplites saurus (Fig. 273B), to relatively deep bodied forms,
i.e., BD 32 to 59% SL in Selene sp. (Aprieto, 1974) (Fig. 274A).
The gut develops as a narrow straight tube on the first day after
hatching. A single gut loop is present in larvae 3-4 mm NL,
which is about 5 days after hatching in Atule mate and Oligo-
plites saurus (Aprieto, 1974; Miller and Sumida, 1974). This
pattern seems to be common among other species although
lengths at which the gut loops vary slightly. The gut extends to
midbody with snout to anus length in preflexion and flexion
larvae usually ranging from 46 to 67% SL (Aprieto, 1974; Lar-
oche et al., MS). The head ranges in length from about 24 to
41% SL and is typically about 33% SL.
Head spines form relatively early in development. The first
head spine to develop is a preopercular spine at the angle of the
518
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 26. Distinguishing Characters Useful in Identification (to Genus) of Flexion and Postflexion larvae of Carangidae. Presence
of character indicated by " + ," absence by "-" and no data by "'0." Species and sources on which this table is based are listed in preceding table.
except for original observations on Gnathanodon speciosus, Naucrates diictor. Parastromaleus niger. and Scomheroides lysan. Character definitions
follow Laroche et al. (MS). Information in this table should be considered preliminary, awaiting more thorough descriptions.
Supra-
occipital
ridge
Angle
preopercular
spine
Supraocular ridge
Posltemporal and
supracieilhral spines
Pterotic
ndgc
Vomer
pigment
Weak
Prom
nent
Dors
Genus
Simple
Ser-
rated
Spinule(s)
Small
spine
Ser-
rated
Ser-
rated
2 or 3
spines
Weak
Promi-
nent
lateral
pigment
Aleclts
+
+
_
-
+
_
—
_
+
_
_
+
_
Alepes
+
+
-
-
+
-
-
-
-)-
-
-
-1-
+
Atropus
0
0
0
0
0
0
0
0
0
0
0
0
0
Atule
-
+
-
-
+
-
-
-
+
-
-
_
+
Campogramma
0
0
0
0
0
0
0
0
0
0
0
0
0
Carangoides
+
+
-
-
+
-
-
-
+
-
-
0
+
Caranx
+
+
-
-
+
-
-
-
-1-
-
—
±
+
"Caranx"
+
+
-
-
+
-
-
—
+
—
_
_
-(-
Chloroscombrus
+
+
-
-
+
-
-
-
+
-
—
+
_
Decaplerus
+
+
-
-
+
-
-
-
-1-
-
-
—
—
Elagatis
+
0
+
-
+
-
-
-
+
-
-
+
+
Gnathanodon
+
+
-
-
+
-
-
-
4-
_
-
-1-
-1-
Hemicaranx
0
0
0
0
+
-
-
-
+
-
_
0
0
Lichia
0
0
0
0
0
0
0
0
0
0
0
0
0
Megalaspis
0
0
0
0
0
0
0
0
0
0
0
0
0
Naucrates
-
+
-
-
-
-
-
-1-
-
+
_
+
+
OUgoplites
-
-
-
+
+
-
-
-
+
-
-
-1-
+
Pamotabus
0
0
0
0
0
0
0
0
0
0
0
0
0
Parastromaleus
+
+
-
-
-
+
-
-
+
-
-t-
Parana
+
+
-
-
-
—
+
—
■¥
_
_
0
-f
Pseudocaranx
0
0
0
0
0
0
0
0
0
0
0
0
0
Scomheroides
+
+
-
-
+
-
-
-
-V
_
—
-1-
■f
Selar
+
+
-
-
+
-
-
-
+
-
_
_
+
Selaroides
0
0
0
0
0
0
0
0
0
0
0
0
0
Selene
+
+
-
-
+
—
-
—
-h
_
—
+
_
Seriola
-
+
-
+
+
-
-
—
+
-
—
+
-1-
Seriolina
0
0
0
0
0
0
0
0
0
0
0
0
0
Trachinotus
-
+
-
-
-
-1-
-
-
-
+
-V
^-
+
Trachurus
+
+
-
-
+
-
-
-
-F
-
-
+
-1-
Ulua
0
0
0
0
0
0
0
0
0
0
0
0
0
Uraspis
0
0
0
0
+
-
-
-
+
-
-
0
0
posterior margin of the preopercle, usually first appearing in
larvae 2.0 to 4.0 mm NL, which coincides with yolk sac ab-
sorption.
Carangids develop two series of preopercular spines, one se-
ries along the posterior margin of the preopercle and another
along the anterior margin, called the "preopercular crest" by
Ahlstrom and Ball (1954). Both series have an upper and lower
segment (Fig. 270). The number of preopercular spines does not
seem to reach a constant number as in larvae of many other
fish families. Instead, the number of spines in both series in-
creases through preflexion, flexion, and postflexion stages to a
maximum of usually about 9 in the anterior and 1 1 in the
posterior series, then decreases in number during transformation
and early juvenile stages. Usually just prior to or during the
early juvenile stage, preopercular spines become completely
overgrown by tissue and bone. Development of preopercular
spines in both the anterior and posterior series proceeds along
the margins away from the angle of the preopercle. Conversely,
reduction in preopercular spination proceeds toward the angle.
When spines are present on either the anterior or posterior
margin, a spine is always present at the angle of the margin, and
it is always the largest. The size and shape of this spine are
particularly useful in distinguishing carangid taxa (Table 126).
For example, Seriola zonata and OUgoplites saurus (Fig. 273C,
273B) have a preopercular spine with a spinule(s), and Elagatis
hipinmilata has serrated preopercular spines (Fig. 273A).
A median supraoccipital crest develops on the head during
the preflexion stage in many species (Table 1 26) and persists
until late in the transformation stage when it becomes overgrown
by tissue and bone. The supraoccipital crest is very useful in
distinguishing carangids since there are relatively few marine
fish families which have larvae with a crest. The shape of the
supraoccipital crest has been used to distinguish carangid larvae
of various taxa (Aboussouan, 1975), however, the difliculty in
defining shape characters makes them somewhat subjective and
of questionable reliability. However, some taxa, i.e., Elagatis
hipinmilata (Fig. 273A) and Chloroscombrus chrysurus (Fig.
270D), have crests which do appear quite distinct from those of
other known carangid larvae.
Among other head spines, supraocular spines and serrations
develop in many taxa (Table 126). The larger multiple supra-
ocular spines present in Naucrates diictor (Fig. 272B) and ser-
rated pterotic ridge in Trachinotus spp. (Fig. 272A) are notable.
All species develop posttemporal and/or supracleithral spines
which vary in number, usually 1-5, and relative size among
taxa (Fig. 270-274).
LAROCHE ET AL.: CARANGIDAE
519
Table 126. Extended.
Vemrolalt
;rai pigment
Aligned
along
myosepla
Internal
melano-
phoi^s
over
dorsal
aona
Lateral
midline
pigment
Melano-
phores
on
branchi-
ostegal
mem-
bi^ne
In-
ternal
melano-
phores
over
noto-
chord
Dorsal
and
anal
(inlet
Melanophi
dorsal body
ares on
margin
Body pigmentation
Body depth
Dorsal
fin
spmes
elongate
{fonn
early)
Pelvic
fin
rays
first to
develop
Dorsal
and anal
fin rays
elongate
(form
early)
Number
Scat-
tered
Anti-
medial
rows
Median
row
Shallow
(A < 35%
SL)
Deep
(i > 35%
SLl
of
myomeres
Dense
Light
(typical)
_
_
_
_
_
_
_
+
_
_
+
-
+
-
-
+
24
+
-
0
+
+
0
-
0
0
+
-
-
+
-
-
-
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
_
+
-
+
+
-
-
-1-
-
-
-1-
-
-1-
-
-
-
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
+
-
0
+
+
0
-
0
0
-
+
-
-1-
-
-
-
24
+
+
+
+
+
+
-
+
+
-
-t-
-
+
-
-
-
24
—
+
0
+
-
0
-
0
0
-
+
-
+
-
-
-
24
-
+
+
+
+
+
-
-
+
-
+
-
+
-
-
-
24
_
—
-
+
-
-
+
+
-
-
+
+
-
-
-
-
25
+
-
-
+
+
-
-
+
-
+
-
+
-
-
-
-
24
+
-
-
+
+
-
-
-
-1-
+
-
-
+
-
-
-
24
0
0
0
0
0
0
-
0
0
0
0
-
+
-
0
-
26
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
+
-
-
+
+
-
-
+
-
+
-
-1-
-
-
-
-
25
+
-
—
+
+
-
-
+
-
+
-
-1-
-
-
-
-
26
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
_
-
-
+
+
-
-
0
0
-
+
-
-f
-
0
-
24
+
-
0
0
0
0
0
4-
-
-
+
0
0
-
-
-
27
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
25
+
-
-
+
+
-
-
+
-
-1-
-
+
-
-
-
-
26
_
+
+
+
+
+
-1-
+
—
—
+
+
—
—
—
-
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
-
-
+
+
-
+
-
-1-
-
-
-1-
-
+
+
+
-
24
_
_
+
+
0
0
-
+
-
-f
-
+
-
-
-
-
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
+
-
+
+
+
-
-
+
-
-1-
-
-1-
-
-
-
-
24
+
-
-
+
-
-
-
-
-1-
-
+
+
-
-
-
-
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24
0
0
0
0
0
0
-
0
0
0
0
-
-1-
-
0
-
24
Dorsal, anal, preanal. and caudal finfolds are present at hatch-
ing. Yolk-sac larvae rapidly develop the pectoral fin base and
finfold. The sequence of fin formation in most species is: caudal,
pectoral, anal and soft dorsal, spinous dorsal, and pelvic. Species
oi Alectis (Fig. 271 A) and Selene (Fig. 274A) are exceptions,
developing pelvic and/or dorsal fin elements precociously before
the notochord begins to flex. The sequence of fin formation in
these taxa is: either pelvic, spinous or soft dorsal followed by
caudal, anal, and pectoral (Aprieto, 1974; Laroche el al., MS).
Spinous dorsal, soft dorsal, and anal fins generally develop
from anterior to posterior, although the first element in each fin
may lag, and the most posterior element in the soft dorsal and
anal fins may develop precociously in some species, i.e., De-
capterus spp. and Selar crumenophthabnus (Laroche et al., MS)
(Fig. 270B, C). In many species at least some dorsal and anal
fin spines ossify from the distal tip proximally (Fig. 27 IC, 272C)
which may be an unusual condition among marine fish larvae
and may help characterize the Carangidae. This condition has
been noted in a number of species and may occur in most or
all species, however, lack of specimens in the critical stage when
this condition is recognizable does not yet permit documenta-
tion of its occurrence. Pectoral fin elements develop from dorsal
to ventral. The pelvic spine develops before the rays, and ray
formation proceeds away from the spine.
The separation of the two anteriormost anal fin spines from
the third spine by a distinct gap is an important characteristic
of most young carangids once fins have formed. This gap is
caused by anterior and posterior extensions of the distal part of
the pterygiophores supporting the second (ultimate) and third
(penultimate) anal fin spines. This gap, although present, is rel-
atively narrow in Elagatis bipinrndala and Seriolina nigrofas-
ciata, which differ from other carangids in having only two anal
fin spines. The only other family known to have young with a
similar gap is the Pomatomidae (Laroche et al., MS).
Development of an "antrorse spine" on the anterodorsal mar-
gin of the first dorsal fin pterygiophore (Fig. 272A) is another
character that is found in most young carangids following fin
development and is shared by only a few other families, i.e.,
Ephippidae. This "spine" is usually covered with skin but is
visible in larvae and juveniles.
Scales begin to develop during the transformation stage. Many
species of carangids develop modified scales in the form of scutes
along the posterior portion of the lateral line. Ossifying scales
are usually first visible along the straight part of the lateral line
anterior and adjacent to the caudal peduncle, where scutes form.
520
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
y*J^-L'
y.^*-j
p^'-A
n'>*^'%A
'^^^
'.^
Fig. 274. (A) Early postflexion larva (5.2 mm) of Selene sp. and (B) late postflexion larva (9.2 mm) of Hemicaran.x ainhlyrhynchus.
Scale development proceeds dorsally, ventrally, and anteriorly
from this location. Berry (1960) presented a detailed account of
scute development and methodology for making counts.
Oi^eo/o^. — Developmental osteology has been described for
Trachurns symmetricus (Ahlstrom and Ball, 1954); Decapterus
punctatus. Elagatis bipinnulata. Selene vomer, and Seriola :on-
a/a (Aprieto, 1974); and .4/M/r ^wa/<' (Miller and Sumida, 1974).
The sequence of ossification is the same for all of these species.
The cleithrum, premaxilla. and posterior preopercular angle spine
are first to ossify in preflexion larvae. Although the cleithrum
begins to ossify early, the pectoral and pelvic girdles do not
completely ossify until late in the transformation stage. Near
the beginning of notochord flexion, the maxilla, dentary, para-
sphenoid, supraoccipital, articular, frontal, angular, and bran-
chial arches begin to ossify. However, much of the cranium does
not completely ossify until late transformation stage. Teeth form
along the anterior margin of the premaxilla as soon as it ossifies.
Aprieto (1974) noted that early ossification of bones related to
feeding is consistent with need for food following yolk resorp-
tion. The first branchial arch begins to ossify first with ossifi-
cation proceeding from the angle of the arch outward. The other
arches ossify similarly in sequence. Gill rakers develop following
ossification of the element on which they are attached. The full
complement of gill rakers is not attained until late transfor-
mation or early juvenile stage. Patches of small teeth form on
the upper pharyngeals of the third and fourth gill arches, and
the fifth arch has tooth patches along most of its length. Pha-
ryngeal teeth ossify early in the postflexion stage.
Vertebrae begin to ossify next, in the middle of the flexion
LAROCHE ET AL.: CARANGIDAE
521
Fig. 275. (A) Postflexion larva (5.5 mm) of Paraslromaleus niger and (B) small juvenile (25.6 mm) of I'raspis secunda.
522
ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM
stage (along with the caudal fin rays) in most species, closely
followed by neural and haemal spines. Vertebrae, neural, and
haemal spines ossify sequentially, anteroposteriorly. Centra os-
sify from their anterior margin posteriorly. Neural spines of the
abdominal vertebrae, and neural and haemal spines of caudal
vertebrae begin to ossify before their respective centra. Ribs
ossify at about the same time and also develop anteroposte-
riorly. Pleural ribs ossify before the epipleural ribs. The urostyle
begins to ossify before the posteriormost two or three vertebrae
during the flexion stage. Ossification proceeds from its anterior
base towards its distal tip as it also does in the hypurals.
Pigmentation. — Details concerning the development and vari-
ety of pigmentation characters are discussed by Laroche et al.
(MS) and are summarized for genera in Table 126. Although
many species have not been observed and this table is tentative,
it reflects the potential utility of pigmentation characters.
It is not possible to describe a generalized pigmentation pat-
tern that is unique to and diagnostic for all carangid larvae. By
the end of the preflexion stage, most species have rows of me-
lanophores along the dorsal and ventral margins of the tail.
Melanophores appear on the head over the brain and eventually
form a cap of pigmentation. Dorso- and ventrolateral pigmen-
tation may be present or absent depending on the species (Fig.
270A, B). A row of small melanophores develops along the
lateral midline at midbody during the preflexion stage and per-
sists into the juvenile stage (Figs. 270-275). When these me-
lanophores are expanded, they appear as a line of pigmentation.
This pigmentation along the lateral midline has been referred
to as the "lateral line streak" by Ahlstrom and Ball (1954) and
Miller and Sumida (1974). The amount and pattern of mela-
nistic pigmentation on the head, body, and fins of carangid
larvae is otherwise quite diverse, grading from very light to very
dark pigmentaton. However, larvae can usually be categorized
as either lightly or darkly pigmented (Table 126, Figs. 270-275).
Darkly pigmented forms usually have a lightly pigmented caudal
peduncle (Figs. 272, 273).
Systematic considerations
Although considerable taxonomic confusion still exists re-
garding carangids, and developmental stages for most species
remain unknown, similarities among larvae of species assigned
to the same genus suggest a congruence between adult and larval
similarities which may reflect the naturalness of some generic
groups. For example, all species of the genus Selene for which
larvae are known share precocious development of the spinous
dorsal, pelvic, and caudal fins, while all species of Decapterus
for which larvae have been described begin development of a
finlet at the posterior of the dorsal and anal fins before more
anterior elements begin to develop. Interestingly, Selar cni-
menophthalmus (which lack finlets as adults) larvae also begin
development of a fin element at the posterior of the dorsal and
anal fins before more anterior elements begin to develop (Fig.
270C). This character may reflect a relationship between De-
capterus and Selar. This type of information is encouraging and
may tend to raise confidence in the naturalness of taxonomic
groups and in the potential utility of developmental characters
for use in systematic studies of carangids.
Developmental information is available for too few species
to allow interpretation of character patterns which might reflect
phylogenetic relationships within the Carangidae. Of course,
investigation of Carangidae's relationship to other groups within
Perciformes is a much larger problem and will require that
similar information be gathered for other taxa. Careful, com-
parative developmental studies are needed to supply this critical
information and provide the most direct route towards a better
understanding of relationships.
(W.A.L.) School of Natural Resources, Department of
Fisheries, Humboldt State University, Arcata, Cal-
ifornia 95521; (W.F.S.-V.) Department of Ichthyology,
The Academy of Natural Sciences, 19th and The
Parkway, Logan Circle, Philadelphia, Pennsylvania,
19103; (S.L.R.) Gulf Coast Research Laboratory, East
Beach Drive, Ocean Springs, Mississippi 39564.
Carangidae: Relationships
W. F. Smith- Vaniz
DESPITE the great economic importance and broad geo-
graphic distribution of the Carangidae, knowledge of their
systematics is very inadequate. The few attempts to determine
their phylogenetic relationships have been both limited in scope
and methodologically flawed. These classifications largely reflect
the distribution of characters shared between taxa rather than
being based on evolutionarily derived characters. Lack of knowl-
edge of an appropriate out-group for comparison has also lim-
ited progress in this area.
In his pioneering study of carangid osteology and relation-
ships, Starks (1911) recognized four subfamilies but stressed the
difliiculty of establishing intrafamilial relationships. Suzuki ( 1 962)
described and illustrated the osteology of 1 8 genera of carangids.
Unfortunately only Japanese species were considered and, al-
though much useful descriptive information was presented, little
progress was made towards attaining a better understanding of
carangid phylogeny. Vergara (1972) described the osteology of
the Cuban species assigned to Caran.x and presented a phyletic
analysis of their relationships. In a subsequent paper Vergara
(1974) expanded his analysis to include all Cuban genera of
Carangidae and evaluated the phenetic relationships of Cuban
Caran.x. Smith-Vaniz and Staiger (1973) concentrated their ef-
forts on the Scomberoidini and presented evidence suggesting
a sister-group relationship between Parana and Scomheroides +
Oligopliles. The detailed comparison and osteological descrip-
tion of Nematistius by Rosenblatt and Bell (1976) provided
SMITH- VANIZ: CARANGIDAE
523
Nematistiidae
(8) (1)(2)
0)
CO '"
(1)
<D
-O CO
CO
— T3
CO
T3
CD
centr
haeni
♦*
c
>< a
CO
0)
n '^
E
^
u L^
o
o
CO o
z
LLI
cr o
Trachinotini
Carang
ni
Scor"*^'"'"i'^i"i '
1
1
1
(95)
-D
(5)
(1)
CD
■o
(4)
en
(20)
(1)
CO
(1)
(1)
CO
E
E
C
CO
(0
(1)
CO
c
o
k_
CO
01
E
o
o
ligoplite
rachinot
(1) ^
.1 i
(9)
«
o
0)
CO
c
o
-
CO
(0
ampogrj
1 other c
Q.
CO
O
H
_l
z
w
w
LLI
o
to
Carangidae
"Carangoi(ds"
Fig. 276. Hypothesized cladogram of "carangoid" fishes including main groups of Carangidae. Numbers opposite rectangles correspond to
characters discussed in text. Numbers in parentheses are estimated total number of species m taxon. Open rectangles are hypothesized to represent
plesiomorphic (phylogenetically primitive) character states and solid rectangles derived character states; characters indicated by barred rectangles
are hypothesized to have evolved more than once, and acquired independently in each lineage so marked; half-barred rectangles indicate that
both the primitive and derived character states occur in some component taxa of the lineage with the derived condition secondarily evolved.
524
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 127. Selected Characters of Carangid Genera. (Abbreviations: Triseg. = trisegmental; Br. = branchiostegal; P = rayless pterygio-
phore.)
1st haemal spine
Caudal peduncle
AM muscle
Tnseg
No. species
Scutes
attachment
grooves
A.'div.
radials
Trachinotini
Trachinotus
(20)
absent
strong
absent
absent
absent
Lichta
(1)
absent
strong
absent
absent
absent
Scomberoidini
Parana
(1)
absent
strong
absent
absent
absent
Scomberoides
(4)
absent
strong
absent
absent
absent
Oligoplites
(5)
absent
strong
absent
absent
absent
Naucratini
Seriola
(9)
absent
weak
present
present
present
Senolina
(1)
absent
weak
present
present
present
Elagatis
(1)
absent
weak
present
present
present
Naucrates
(1)
absent
weak
present
present
present
Campogramma
(1)
absent
weak
present, but
rudimentary
present
absent
Carangini
Alectis
(3)
present
strong
absent
present
absent
Alepes
(4)
present
strong
absent
present
absent
Atropus
(1)
present
strong
absent
present
absent
Atule
(1)
present
strong
absent
present
absent
Carangoides
(22)
present
strong
absent
present
absent
Caranx
(14)
present
strong
absent
present
absent
Chloroscombrus
(2)
present
strong
absent
present
absent
Decapterus
(10)
present
strong
absent
present
absent
Gnathanodon
Hemicaranx
Megalaspis
Pantolabus
(1)
present
strong
absent
(4)
present
strong
absent
(1)
present
strong
absent
(1)
present
strong
absent
present
absent
present
absent
present
absent
present
absent
SMITH- VANIZ: CARANGIDAE
525
Table 127. Extended.
Ir rays
Infenor vertebral
foramina
Epural bones
Vertebrae
'Predorsal formulae
7-8
8
absent
present
3
3
10 + 14
10 + 14
O/O/O + P/l + l/
0/0/0 + P/P+1/
0/0/0+1/
V-VI + I, 17-29
II + I, 16-18
VII + 1, 19-21
II + I, 17-19
9
absent
3
10 + 17
8
absent
2
10 + 16
7
absent
2
10 + 16
0/0/0+1/
0/0/0+1/
0/0/0 +P/1/
O/O/O + P/l + l/
0/0/0 + P/P+1/
0/0/0 + P/P/1/
0/0/0 + P/P+P/1/
VI + 1. 32-38
II + I, 34-38
Vl-VIl + I, 19-21
II + I, 16-20
IV-VI + I. 18-21
II + I, 19-21
absent
absent
absent
absent
absent
10 +
14(6)
11 +
13(2)
11 +
14(1)
11 +
13
10 +
14
10 +
15
10 +
14
0/0/0+1 + 1/
0/0/0 + 2+1/
0/0 + 0/1 + 1 + 1/
0/0/0/0+1/
0/0 + 0/1 + 1/
0/0 + 0/1 + 1 + 1/
0/0/0/1 + 1/
0/0/0+1/
0/0 + 0/1 + 1/
VII-VIII + 1,22-39
7
present
2
10 +
14(2)
0/0 + 0/1 + 1/
10 +
15-16(1)
0/0+0/P+l/
7
present
2
10 +
14
0/0+0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
7
present or
2
10 +
14(21)
0/0 + 0/2+1/
absent (1)
10 +
15(1)
7
present
2
10 +
10 +
14(11)
15(3)
0/0 + 0/2+1/
7
present
2
10 +
14
0/0/0+1 + 1/
0/0+0/2+1/
7
present
2
10 +
14(9)
0/0 + 0/2+1/
10 +
15(1)
0/0/0/2+1/
-/0/0/2+1/
-/0/0+0 + 2+1/
7
present
2
10 +
14
0/0 + 0/1 + 1/
7
present
2
10 +
15(2)
0/0 + 0/1 + 1/
10 +
16(2)
0/0 + 0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
II + I,
15-
■22
VII + I,
30-
■37
I+I,
15-
-18
VI + I,
25-
-30
I+I,
18-
-20
IV-V + I,
25-
-29
II+I,
15-
-17
VI-VII + I.
26-
-28
II+I,
23-
-25
VI-VII + I,
18-
-22
II+I,
16-
-20
VIII + 1,
23-
-27
II+I,
18-
-23
VIII + I,
19-
-22
II+I,
17-
-18
VIII + I,
22-
-25
II+I,
18-
-21
VIII + I.
18-
-35
II+I,
16-
-29
VIII + I,
18-
-25
II+I,
14-
-21
VII-VIII + I,
25-
-28
II+I,
25-
-28
VIII + I,
,27-
-37
11 + 1,22-32
VII + I. 18-21
II + I, 15-17
VI-VII + I. 20-25
11 + 1,20-25
VIII + I, 18-20
II + I, 16-17
VIII + I, 21-23
II + I. 18-20
526
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 127. Continued.
No, species
Scutes
1st haemal spine
attachment
Caudal peduncle
grooves
AM muscle
Ai'div.
Tnseg.
radials
Parastromaieus
(1)
present
strong
absent
present
absent
Pseudocaranx
(3)
present
strong
absent
present
absent
Selar
(2)
present
strong
absent
present
absent
Selaroides
(1)
present
strong
absent
present
absent
Selene
(7)
present
strong
absent
present
absent
Trachurus
(12)
present
strong
absent
present
absent
Ulua
(2)
present
strong
absent
present
absent
Uraspis
(3)
present
strong
absent
present
absent
definitive evidence supporting its removal from the Carangidae.
Their work is an important contribution towards elucidating
carangoid relationships.
Much more effort needs to be focused on obtaining basic data
on the biology, ontogeny and systematics of carangids. The data
presented in Table 127 and the following discussion are a first
step in that direction.
Relationships
Evidence supporting the monophyly of carangoids and several
major groups of carangids is discussed below. Refer to G. D.
Johnson (this volume) for discussion of interfamilial relation-
ships of the three echeneoid families. The oldest available name
for each of the four carangid tribes herein recognized has not
been determined but none is proposed as new. No synapomor-
phies were found to support the inclusion of Lichia in the tribe
Trachinotini, and its placement is one of convenience in accord
with the practice of previous authors and reflects my own sub-
jective bias. Autapomorphies that define Trachinolus. Lichia
and the naucratine genera are not included in the cladogram
(Fig. 276) because they are not informative about relationships.
These taxa are recognized individually in the figure to make it
easier for the reader to determine the character state distribu-
tions and the number of species comprising each genus. The
Carangini includes approximately 20 genera (Table 127), many
not well established, and their osteology poorly known. Until
this presumably monophyletic assemblage has been studied in
much greater detail no meaningful discussion of relationships
will be possible. Several recent authors have considered Para-
stromateus to constitute either a monotypic family or carangid
subfamily. Available evidence suggests, however, that it should
be assigned to the Carangini.
The following character states are the basis for the hypotheses
of carangoid interrelationships inferred in Fig. 276. The pre-
sumed derived character state is listed first, followed by dis-
cussion of the character in out-groups when necessary.
(I) Freihofer (1978) made the important observation that in
the Nematistiidae, Carangidae, Coryphaenidae, Rachycentridae
and Echeneididae there are one or two tubular ossifications
(prenasals) around the anterior extension of the nasal canal. This
presumed specialization of the lateralis system is very rare in
percoids (also present in the unrelated and highly specialized
Toxotidae) and is considered to be a synapomorphy suggesting
that these five families constitute a monophyletic group.
(2) The possession of small adherent cycloid scales is a derived
character shared by carangoids in contrast to the typically cte-
noid scales of most other percoids. Berry (1969) reported that
the carangid Elagatis has "ctenoid" scales and Zheng (1981)
also described the highly modified caudal peduncle scales of
Naucrates as ctenoid. These scales are not typically "ctenoid"
and appear to represent modifications of the carangoid scale-
type.
(3) Two separate prenasal canal units, one membranous and
one bony (Carangidae) or both bony (echeneoids). In contrast,
Nematistiits has only a single prenasal canal unit.
(4) Loss of the bony stay (Fig. 277) posterior to ultimate dorsal
and anal pterygiophores that is present in most other percoids
(see Table 127, G. D. Johnson, this volume).
(5) On shoulder girdle, middle part of coracoid with its an-
terior margin consisting of a lamella of bone broadly extending
towards the median cleithral wing (Suzuki, 1962: figs. 36-44).
In Nematistius the middle and lower parts of the coracoid are
rodlike with lamellar bone restricted to its posterior margin
(Rosenblatt and Bell, 1976; fig. 8).
(6) Basioccipital with a pair of foramina (Rosenblatt and Bell,
1976: fig. 3) into which anterior processes of the gas bladder
extend forward to the region of the inner ear.
(7) Anterior shift of second predorsal bone to the first inter-
neural space and first pterygiophore greatly expanded and plate-
like. In carangids, as in most other percoids, the second pre-
dorsal bone occupies the second iniemeural space (predorsals
absent in echeneoids), and in both echeneoids and carangids the
first dorsal pterygiophore is not greatly expanded.
(8) Spines of first dorsal fin very long and filamentous and
only basally connected by interradial membrane.
(9) Tubular ossifications surrounding both prenasal canal units;
SMITH- VANIZ: CARANGIDAE
Table 127. Continued. E.xtended.
527
Br. rays
Infenor vertebral
foramina
Epural bones
'Predorsal formulae
Anal tin
7
present
2
10 +
14
0/0+0/1 + 1 + 1/
7
absent
2
10 +
14(2)
0/0+0/2+1/
10 +
14-
15(1)
7
present
2
10 +
14
0/0 + 0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
7
present
2
10 +
14
0/0+0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1/
7
present
2
10 +
14
0/0 + 0/2+1 + 1/
IV-V,41-44
II + I,
VIII + I,
35-39
23-28
II + I,
VIII + I,
20-24
23-27
II + I,
VIII + I,
19-22
24-26
II + I,
VIII + I,
21-23
20-24
II + I,
VIII + I,
16-20
28-36
II + I,
VIII + I,
24-32
21-22
II + I,
VIII + I.
17-18
24-32
II + I, 17-28
' Methodology of predorsal formulae after Ahlslrom el al. (1976).
posterior canal unit unossified in carangids and absent in Ne-
matistiidae.
(10) Absence of the so-called beryciform foramen in the an-
terior ceratohyai.
(11) Absence of predorsal bones.
(12) Several anal pterygiophores anterior to the first haemal
spine (versus one in Carangidae, Nematistiidae and most per-
coids).
( 1 3) Larvae very elongate, with dorsal fin ray development
not completed until two or three times size at flexion (G. D.
Johnson, this volume). In contrast, larvae of carangids are mod-
erate to deep-bodied and complete dorsal and anal fin devel-
opment in conjunction with or soon after flexion (Laroche et
al.. this volume). Larvae of Nematistiidae are unknown.
( 1 4) Posteroventral elongation of first proximal pterygiophore
of anal fin resulting in a relatively wide gap (Fig. 278b-e) be-
tween the last two anal spines. The carangid genera Elagatis
and Seriolina (Fig. 278c) are exceptional in having only one
spine on this pterygiophore so the gap is not as apparent.
(15) Presence of a separate dorsal division (A,') of the ad-
ductor mandibulae muscle originating on the suborbital shelf
(Fig. 279). The relative size of the suborbital shelf in carangids
is not correlated with the presence or absence of this muscle,
which is also lacking in echeneoids, Nematistius and most per-
coids.
(16) Some lateral line scales (at least those on caudal peduncle)
modified as thickened scutes.
( 1 7) Caudal-peduncle grooves present dorsally and ventrally;
these specialized structures undoubtedly have a hydrodynamic
function related to swimming mode. Campogramma. which
appears to be the most advanced naucratine (judging from the
relatively large number of autapomorphic characters that it pos-
sesses), is exceptional in having only rudimentary caudal-pe-
duncle grooves (absent in young).
The occurrence of caudal-peduncle grooves on Nematistius
which shares many plesiomorphic characters, including a sim-
ilar external morphology, with naucratines is most parsimoni-
ously interpreted as parallelism. These structures are also pres-
ent on carcharhinid sharks.
(18) Premaxilla non-protractile and in adults dorsal margin
of upper lip attached to snout by a broad frenum.
(19) Epiotics broadly united along midline of cranium beneath
the supraoccipital.
(20) Total vertebrae 26 or 27 (versus 24 or 25).
(21) Cheeks unsealed.
(22) Spines of dorsal and anal fins with well developed venom
glands (Halstead et al, 1972; Sazima and Uieda, 1979).
(23) First proximal pterygiophore of anal fin expanded an-
terolaterally to form a roof over anal spines (Smith-Vaniz and
Staiger, 1973: fig. 15b).
(24) Juveniles with two widely spaced rows of dentary teeth
Fig. 277. Terminal pair of dorsal fin rays and associated pterygio-
phores: (a) Nemalislius pecloralis (Note presence of large bony stay
behind last medial pterygiophore) and (b) Naucrates duclor.
528
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
Nematistius
Caranx
Parastromateus
Fig. 278. Anterior pterygiophores and associated spines and rays of anal fin (Note relative spacing between last two spines); (a) Nemalislius
pectoralis; (b) Senola zonala; (c) Seriolina nigrofasciata; (d) Caranx sexfasciatus; (e) Parastromateus niger.
into which the premaxiliary teeth fit when the mouth is closed,
and the outer series of dentary teeth strongly hooked outward
and with spatulate tips. Major ( 1 973) has shown that this dental
arrangement facilitates lepidophagous feeding in juvenile Sconi-
heroides and, on the basis of stomach content analyses of two
species, concluded that at least some Oligoplites have similar
feeding habits. Carr and Adams (1972) postulated that inten-
tional removal of ectoparasites is also an important activity in
juvenile Oligoplites. Presumably such unique dentition facili-
tates both types of specialized trophic ecology.
(25) Interosseous space between coracoid process of dentary
and posterodorsal projection of anguloarticular minute or ab-
sent.
(26) Pleural ribs on vertebrae 3 through 7 or 8 attached high
on centrum and spatulate in cross-section.
(27) Posterior dorsal- and anal-fin rays consisting of semi-
detached finlets.
(28) Reduction in number of epurals in caudal fin from 3 to 2.
(29) Supramaxilla minute or absent. It might be argued that
the reductive-loss supramaxilla character state is a synapomor-
phy uniting Trachtnotits + Lichia with the Scomberoidini, in
which case the well developed supramaxilla of Parana would
constitute a reversal. Alternatively, the reductive trend of the
supramaxilla in the two taxonomic pairs under consideration
might be a simple case of parallelism. In the absence of any
other obvious synapomorphy that supports the first hypothesis
SMITH-VANIZ: CARANGIDAE
529
Fig. 279. Adductor mandibulae: (a) Seriola diiinerili; (b) Caran.x sexfasciatus. Note the presence of A,', a separate dorsal section originating
on the suborbital shelf (S).
and because the reversal of a reductive trend is involved, I
believe it is more conservative (even though less parsimonious)
to retain the unresolved position of Trachinolus-Lichia in the
cladogram.
(30) Pelvic fins absent at all stages of development.
(31) Increase in number of caudal vertebrae from 16 to 17.
(32) Branchiostegal rays 9 (versus 7 or 8).
(33) Basibranchial dentition consisting of large median tooth
plates, presumably derived from fusion of the large paired tooth
plates found in Scomhcroides and Oligoplitcs (Smith-Vaniz and
Staiger. 1973: figs. 24b-d).
(34) Lateral line with 5-9 dorsal branches.
(35) Loss of dorsal-fin spines resulting in an increase in the
number of rayless pterygiophores (see Table 127).
(36) Loss of mesopterygoid teeth.
(37) Loss of supramaxilla (minute in Scninberoides).
(38) Loss of suborbital shelf on third infraorbital bone.
(39) Infraorbitals 2-4 enlarged and extending posteriorly across
cheek in adults.
(40) Prominent dark spots or short bars on sides of adults.
Unlike many carangids, the juveniles of both Scomberoides and
Oligoplites are unbarred.
Recognition of the family Nematistiidae
The familial placement oi Nematistius has long been contro-
versial. Some distinguished ichthyologists (Gill. 1863; Jordan
and Evermann, 1896-1900; Berg. 1947) placed it in a separate
family while others, most recently Robins et al. ( 1 980). assigned
it to the Carangidae. On the basis of a detailed osteological
comparison. Rosenblatt and Bell (1976) concluded that Ne-
matistius should not be classified with the Carangidae. They
also commented on the striking similarities between the Ne-
matistiidae and certain primitive carangids, especially naucra-
tine genera. Almost all of the many features shared by these two
530
ONTOGENY AND SYSTEMATICS OF nSHES-AHLSTROM SYMPOSIUM
taxa are plesiomorphic character states, the one notable excep-
tion being caudal-peduncle grooves.
In addition to possessing different character states 3-8 and
14 as listed above, Freihofer (1963) observed that the Caran-
gidae and Nematistiidae differ in the course of the nerves of the
ramus lateralis accessorius (RLA) complex; the former having
pattern 9 and the latter pattern 10 (reduced). Nematistius also
differs in having two foramina in the scapula; a typically large
one and a smaller more posteriorly positioned foramen (absent
in carangids) that also occurs in the Rachycentridae. Like the
two RLA nerve patterns, the derived character state for the two
scapular foramina conditions has not been determined. Never-
theless, the inclusion of Nematistius in the Carangidae would
make the family paraphyletic (unless the three echeneoid fam-
ilies are also included) and impossible to define based on shared
derived characters.
Familial position of Parastromateus
Several recent authors have followed Apsangikar (1953) or
Suzuki ( 1 962) in recognizing Parastromateus either as a subfam-
ily of the Carangidae or as the sole representative of the mono-
typic Formionidae (=Apolectidae or Parastromatidae). All the
characters used to justify the latter classification, with one ex-
ception discussed below, have been autapomorphic characters
which can provide no information about relationships. That the
genus should be assigned to the Carangidae is clearly indicated
by the possession of derived character states 3-5 and 14-16
discussed previously.
Haedrich (1971) noted that Parastromateus (=Apolectus) is
the only fish with a pattem-9 ramus lateralis accessorius nerve
system that has a "pons moultoni." In an addendum to his
paper it was suggested that retention of the pons is a primitive
character state. It should be emphasized that very few carangoids
have been examined for the presence of this easily overlooked
structure. Until the distribution of this character has been de-
termined for the major lineages of carangoids, its phylogenetic
significance can not be evaluated. Similarly, no data have been
presented to substantiate assigning Parastromateus to its own
subfamily within the Carangidae.
Department of Ichthyology, The Academy of Natural Sci-
ences, 19th and The Parkway, Logan Circle, Phila-
delphia, Pennsylvania 19103.
Mugiloidei: Development and Relationships
D. P. DE Sylva
MUGILOIDEI is one of three closely related suborders, to-
gether with Sphyraenoidei and Polynemoidei, in the Per-
ciformes. The suborder is represented by a single family, the
Mugilidae. Until recently, the Atherinidae had been considered
close relatives of the Mugiloidei. Within the family Mugilidae,
classical morphological taxonomic analyses have been applied
to regional groupings rather than to the family as a whole (Weber
and de Beaufort, 1922;Roxas, 1934; Smith, 1935, 1947;Schultz,
1946; Ishiyama, 1951; Thomson, 1954; Ebeling, 1957, 1961;
Lindberg and Legeza, 1969; Ben-Tuvia, 1975). Hence, the sys-
tematics of the family are poorly understood.
Mullets are characterized by thick, streamlined bodies, deeply
forked caudal fin. large cycloid or weakly ctenoid scales, and
the lack of a lateral line. The mouth is small, the jaws have
small teeth or none, and the gill rakers are long and slender, the
latter assisting the pharyngeal jaw apparatus to form a filtering
apparatus (Lauder and Liem, 1983). They share, with the thread-
fins and barracudas, the characteristic of having two widely
separated dorsal fins. Two subfamilies of mullets are recognized,
the Mugilinae and the Agonostominae (Jordan and Evermann,
1896-1900). The latter have sessile teeth which attach directly
to the jaws, a flat preorbital, and only 2 anal spines in the adult.
The Mugilinae have flat labial teeth, if any, connected to the
jaws by elongated fibers, a ridged and grooved preorbital, and
3 anal spines in the adult.
The Mugilinae occur worldwide except in polar regions, while
the Agonostominae are confined to Central America, the west-
em Indian Ocean, the tropical west Pacific, and the Australian
coastline. Mullets occur in oceans, bays, estuaries, and fresh
water. They are uniformly important as food for humans and
an important prey in the food web. They seldom exceed 1 meter.
Development
Many studies exist on the eggs, larvae, and post-larval stages
of mullets in comparison to other families, but only a few are
comprehensive, and most deal with a single species (e.g., An-
derson, 1957;Dekhnik, 1973; Farrugio, 1977; Kuo et al.. 1973;
Lai. 1979; Martm and Drewry, 1978; Sanzo, 1936; Tung, 1973;
Vialli, 1937; Yang and Kim, 1962; Yashouv and Bemer-Sam-
sonov, 1970). However, a general overview of each stage can
be summarized.
Eggs are pelagic, spherical, and transparent, with the surface
of the egg being smooth and usually without sculpture (Fig. 280).
The yolk is unsegmented. the perivitelline space is narrow, and
there is one or more oil globules. During development, several
oil globules merge with each other, becoming situated on the
yolk sac upon hatching. Egg sizes for various species of European
and African mugilids range from 0.6 to 1 .3 mm and vary greatly
in diameter from one geographic area to another. Although most
eggs have similar pigmentation, different species have similar,
though sometimes overlapping spawning seasons, which may
offer a clue in the analysis of phyletic relationships and mugilid
evolution.
Larval pigmentation ranges from relatively light to heavy (Fig.
DE SYLVA: MUGILOIDEI
531
Fig. 280. Various stages of development of eggs of silver mullet. Mugil curema: (a) unfertilized eggs; (b) 2 hours after fertilization (32 blastomeres);
(c) 4 hours after fertilization (blastodisc well formed, cells small); (d) 8 hours after fertilization (segmentation cavity forming); (e) 12 hours after
fertilization (early embryo); (0 16 hours after fertilization (embryo); (g) 24 hours after fertilization (lateral view of embryo); (h) 24 hours after
fertilization (top view of embryo); (i) 32 hours after fertilization (lateral view of embryo) (from Anderson, 1957).
281). All larvae have stellate melanophores on the oil globule,
which also occur on the forehead of some species. This feature
has not been studied for mugilids on a global basis, but offers
possibilities for phyletic analysis.
At hatching, stellate melanophores also occur on the yolk
surface and body, with fine spots along the dorsal and ventral
profile of the caudal trunk. The caudal rays form first, at 4 mm
total length. The second dorsal forms at between 4 and 5.7 mm.
and the first dorsal forms at 5.4 mm. Scales begin to develop
at between 8 and 10 mm and are well formed at 1 1 mm. Pig-
mentation is strong at from 2 to 5 mm, and the dorsal surface
is dark by 5 mm total length. By a length of 8.2 to 10.9 mm
they are silvery white to silvery green, and at this size they
resemble the adults in body form, there being no distinctive
metamorphosis throughout development (Fig. 281).
Identification of later larvae (Fig. 282) is based upon color.
532
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 281. Larvae of silver mullet. Mugil curema. (A) Newly hatched, 1.76 mm; (B) yolk-sac stage, 2.15 mm; (C) yolk-sac stage, 2.47 mm; (D)
yolk-sac stage, 2.56 mm; (E) yolk-sac stage, 2.56 mm; {¥) 3.7 mm; (G) 4.0 mm; (H) 4.7 mm; (I) 5.3 mm. From Anderson (1957).
pigmentation pattern, number of anal elements, longitudinal
scales, transverse scales, scale morphology, pyloric caeca, and
gill rakers. The general profiles of the head, lips, and the labial
and lingual teeth are also very useful characters (Tung, 1973;
Wallace and van der Elst, 1973; Thomson, 1975; Lai, 1979;
Zisman, 1982).
Relationships
In some species such as Mugil cephalus, as presently under-
stood, which has a worldwide distribution, there is considerable
variability in meristic characters and proportional measure-
ments. Additional studies are warranted to determine the real
extent of genetic exchange between local subunits (Thomson,
1982).
At the generic and specific levels, mugilid taxonomy has not
been resolved. As in the case of Mugil cephalus, those species
with extensive ranges may be known under different names in
various parts of their range.
A variety of external morphology features have been used to
identify genera and species of the adult stages, ranging from
dentition (Ebeling, 1957, 1961; Farrugio. 1977) and scales
(Thomson, 1982), to eye coloration (Alvarez-Lajonchere, 1975).
Internal anatomy is valuable in systematic analysis, including
the shape and number of pyloric caeca (Perlmutter et al., 1957;
Luther, 1975b), the alimentary tract (Thomson, 1966), intestinal
convolution (Hotta, 1955), osteology (Luther, 1975a; Mohsin,
1978; Kobelkowsky and Resendez, 1972; Sunny, 1971; Hotta
and Tung, 1972), and otoliths (Morovic, 1953).
Phyletic studies within the family have not been undertaken.
Thomson's manuscript revision (see Thomson, 1982) recog-
nizes 14 genera and 64 species of the nominal 282 species. Of
these, 32 are indeterminate because of inadequate descriptions
or missing holotypes. The only published world revision, by
Schultz (1946), recognizes 13 genera. Relationships are based
upon the adipose eyelid, type of scales, labial characteristics,
preorbital shape, and type of habitat. Larval mullets have been
studied extensively, but not on a worldwide basis, and no phy-
letic analysis has been attempted. It is known that in certain
species the young stages have 2 anal spines, but larger stages
have 3 spines. The younger stages have been referred to as the
DE SYLVA: MUGILOIDEI
533
Fig. 282. Postflexion larvae (A, B) and juvenile (C) of silver mullet.
Mugil curema. (A) 7.0 mm; (B) 14.5 mm; (C) 25.5 mm. From Anderson
(1957).
"querimana stage." An analysis of the genera and species pos-
sessing this trait has not been undertaken. Biochemical studies
on mugilid systematics have been undertaken by Callegarini
and Basaglia (1978) and by Autem and Bonhomme (1980) in
the Mediterranean, but no studies have been carried out on a
worldwide basis.
As stated in the discussions on Sphyraenoidei and Polyne-
moidei (this volume), they have been closely linked with the
Mugiloidei phyletically. Previously, the athennids had been
placed within this assemblage, but Rosen ( 1 964) has clearly
shown that the atherinids belong in a separate superorder con-
taining the flyingfishes and livebearers. The Mugiloidei appear
more closely related osteologically to the Sphyraenoidei than
they are to the Polynemoidei.
A brief history of the higher classification of these groups is
reviewed here. The suborder Percesoces had included the Ath-
erinidae. Mugilidae, and Sphyraenidae (Jordan and Evermann,
1868-1900), but Starks (1900) questioned their similanty, though
he believed them to be quite close based upon the decided
branching of the epiotic crests. Superficially, the mugiloid-sphy-
raenoid skeleton resembles that of atherinoids, but Hollister
(1937) pointed out an important developmental difference be-
tween them. In Athehna. the lowermost hypural plate develops
as a single entity. In Mugil and Sphyraena this plate forms from
two distinct elements. Berg ( 1 940) separated the Mugilidae, with
the Sphyraenidae and the Atherinidae, from the Perciformes as
the order Mugiliformes because they have abdominal pelvic fins,
a relatively primitive character. Rosen (1964) also pointed out
the similarities among mugiloids, sphyraenoids, and polyne-
moids in ossification of the skull, especially the common pres-
ence of a subocular shelf, the jaw suspension and feeding mech-
anism, jaw musculature, and the pharyngobranchial and
opercular apparatuses. Further, Rosen stated that "the embryos
of mullet (Anderson, 1957) and barracuda (Orton, 1955b) are
small and contain a large oil globule .... A forward-displaced
heart is also characteristic of Oryzias . . . but not of Sphyraena
(Orton, 1955b;Shojimaetal., 1957), and probably not of A/i/.g//."
Removal of the Mugilidae from the suborder Percoidei is
supported by studies of blood plasma and plasma proteins (Sul-
ya et al., 1960). Plasma proteins of mugilids are less complex
than those of any other family considered to be Perciformes,
and show relationships to some species of Cypriniformes and
Clupeiformes (Gunter et al.. 1961). In contrast, plasma proteins
of some species of 3 perciform families, the Carangidae, Sciaeni-
dae, and Scombridae, do not differ greatly from those of the
Mugilidae. Based on this, the Mugilidae could be regarded either
as belonging to the most primitive perciform group or as branch-
ing from some early perciform.
The early life history stages do not appear to offer useful hints
as to phyletic relations with other taxa, except that the Mugi-
loidei have 23 myotomes dunng larval development, a feature
shared with the Polynemoidei and Sphyraenoidei.
RosENSTiEL School of Marine and Atmospheric Science,
University of Miami, 4600 Rickenbacker Causeway,
Miami, Florida 33149.
Sphyraenoidei: Development and Relationships
D. P. DE Sylva
SPHYRAENIDAE is a closely knit, monogeneric perciform
family of the suborder Sphyraenoidei (Gosline, 1971),
Schultz (1953) revised the family, which has since been partially
modified by Smith ( 1 956b), Williams ( 1 959), and de Sylva (1975)
for Indian Ocean species. Six genera, including three new names,
were proposed by J. L. B. Smith in his 1 956 review of the Indian
Ocean species. These have been synonymized by subsequent
authors to include the single genus Sphyraena, recognized for
all living species. Fossil genera have been noted in the Creta-
ceous and are widespread since the lower Eocene. These are
represented by the genera Sphyraenodus, Protosphyraena, Pro-
sphyraena. and Sphyraena (see de Sylva, 1963). However, be-
cause most fossil generic descriptions are based only upon teeth
or dentary fragments, it seems presumptive to attach very great
importance to the validation of such genera. In a draft revision
of the family, I have recognized the genera named by Smith, as
well as other genera previously proposed for other sphyraenids,
at the subgeneric level to clarify phyletic relationships on a
worldwide basis (Fig. 283; Table 128).
All species are tropical or temperate, and are schooling or
solitary predators. They usually live in the littoral zone from
the surface to just off the bottom in shelf waters. Several are
epipelagic and are found far from shoal water. They are im-
portant food fishes, although one species, Sphyraena barracuda.
is frequently responsible for ciguatera poisoning (de Sylva, 1 963).
Maximum size is 180 cm and 48 kg.
There are 20 valid species of the 69 nominal species. Sphy-
raenids are distinguishable from Polynemidae and Mugilidae
by their well-developed fang-like teeth, large mouth, and point-
ed snout, with the upper jaw not protrusible. Gill rakers may
be absent, bristle-like, or limited to one or two at the angle of
the gill arch (de Sylva, 1975).
Development
Eggs of Sphyraenidae have been described for only 3 species,
and they are similar in size and pigmentation. Larval stages
have been described for 5 (Raffaele, 1 888; Bamhart, 1927; Vial-
h, 1956; Orton, 1955b; Shojima et al., 1957; Mannaro, 1971;
Uchida et al., 1958; de Sylva, 1963; Houde, 1972b). Larval
stages have been described for 4 of the 20 species, from rea-
sonably complete developmental series (e.g.. Figs. 284-287).
Osteological development of the neurocranium is described for
only 1 species (Gregory, 1933), while the caudal skeleton and
urophore complex have been studied for only 3 species (Hoi-
CO
•H
u u
u
■H QJ
o
C i-^
O -H
(0
Q. QJ
<U
3
M
CO ^
CO
^-^
'i
J^
nj
•1—)
■H
■H
w
CD
nJ
-a
J-i
3
nj
DC
•H
w
c
QJ
C
c
TO
c
CO
■H
CO
CO
to
O
^
0
OO
(d
T)
QJ
4J
c
c
rH
CO
•H
CO
x:
OJ
a>
4J
CO
>,.
-hI
3
CO
CO
c
CO
rH
CO
w
nJ
m .H
m
3
(J
to
(U
to
-3
tn
x:
cd
4-1
W
■H
u
o
c
OJ
QJ
•H rH
3
(U
3
•H
T-\
CJ
3
•H
u
u
o
3
p.
■l-\
x:
to
4-1
u
rH -H
CU
to
to
e
X3
CO
u
3
nJ
•H
tfi
4J
••-i
s^
(U
en
cn
C
CO -a
T3
V*
to
x:
2
O
CU
•H
CO
ftO
1-1
>
>,
J^
4-1
14-1
CO
to
CO
QJ
<U 3
•H
>1
•H
o
3
i-H
•H
CO Q-
U
p
Xi
CO
u
o
3
CO
>
u
■H
an
U U
U
x:
CO
to
■u
f-A
3
J-i ^— '
u
■H
r-H
^
u
0
3
■o
u
O -H
•H dJ
3
3
3
QJ
S^
<4-l
CO
a.
1
^
o
I
(0
1
1 1
a
XI Q.
> tn
at
m| a.[
■r-l
CO
X>
\
\
(subgen.
nov)
/
\
(Australuzza)
/
(Sphyraenella)
3D, 4D, 5B. \
7E, 8A, \
2a\3C, 4A,
5A,
(Sphyra
7G, \
2na)
3A, AA, 5B,
7D, 8C
\
/
(Agrioposphyraena)
\
2B, 3A, 5B,
!, 2B
5A,
7E?\
y
^
8B
6,
\
\
3A,
\
\
V
\
\
8A,
9A,
lOA
\
\
\
\
y
Sph
yraena forster
i
\
X
y
7A, 8C,
lOA
(
\
Callosphyraena)
y
S
ph
yraeni
dae
Fig. 283. Diagram of relationships among sphyraenids based on adult and larval characters. Numbers refer to characters listed in Table 128.
Labelled horizontal lines cross branches and demonstrate presumed advanced character states.
534
DE SYLVA: SPHYRAENOIDEI
535
i ^-"4
Fig. 284. Developmental stages of Sphyraena borealis reared in the
laboratory. (A) 3.8 mm; (B) 4.3 mm; (C) 5.3 mm; (D) 7.4 mm (from
Houde, 1972b).
VJ.V-
Fig. 285. Developmental stages of Sphyraena borealis. Specimens
A, B, and C were laboratory reared; specimen D was collected in a
plankton net. (A) 9.4 mm SL; (B) 12.3 mm SL; (C) 14.5 mm SL; (D)
21.0 mm SL(from Houde, 1972b).
Fig. 286. Drawings showing changes in pigmentation and body form
with larval development in Sphyraena barracuda. (A) 5.5 mm SL; col-
lected by R/S DANA, Station 1293-V, 17°43'N, 64°56'W, April 17,
1922. (B) 6.6 mm SL; collected by R/S DANA, Station 952, 17°55'N,
64°48'W, May 12, 1921. (C) 8.6 mm SL; collected by R/S DANA,
Station 1352-V, 35°42'N, 73°43W, May 21, 1922. (D) 11.9 mm SL;
collected by Donald P. deSylva, 1 mile southwest of the harbor entrance
of North Bimini, Bahamas, June 6, 1956 (from de Sylva, 1963).
lister, 1937; Monod, 1968). Development of Sphyraena is di-
rect, with no metamorphosis (Vialli, 1956; de Sylva, 1963;
Houde, 1972b).
Meristic characters are not especially valuable in differen-
tiating most adult species of this family. Although little work
has been done on larval meristic characters, it would be expected
similarly that they would not prove valuable. Anal rays vary
from 8 to 9, and the dorsal secondary rays of the caudal fin vary
from 9 to 10 in two different subgenera.
Similarly, morphological characteristics do not differ widely
in the early life history of the species except that two groups
can be broadly identified— those with blunt heads and more
fusiform bodies, such as S. barracuda (de Sylva, 1963) (Figs.
286, 287) and those with more slender heads and having fleshy
tips on the lower jaw and a more slender, tapering body, as in
S. sphyraena and S. borealis (ViaWi, 1956; Houde, 1972b) (Figs.
284, 285).
Adult species are distinguished by the shape and angle of the
teeth, number of lateral line scales, opercular and preopercular
bone configuration, lateral pigment pattern, dorsal fin place-
ment, and kinds of gill rakers.
In 5. barracuda, adult characters are acquired over a size range
of from 5.5 to 2 1 3 mm. Pigmentation is acquired gradually from
about 5.5 mm to 24 mm, then rapidly above that size.
In S. barracuda, the caudal fin forms first followed by the
536
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 128. Characteristics of Sphyraenidae. (+ = occurs in this species;
■ does not occur in this species; 0 = no information.)
Sphyracna
forslen
S
pingiiis
S.
Ilavicauda
S chryso-
taenia
S
heilen
S aculi-
pinnis
S novae- S. S. S.
hollandiae tucasana idtasles argentea
1. Meristic
Lat. Hne scales
2. Maximum length (mm), SL
3. Gillrakers
a. Absent
b. Occur as spinules
c. One
d. Two
4. Lower jaw
a. With fleshy knob
b. Without fleshy knob
5. Dorsal fin
a. Above pelvics
b. Behind pelvics
6. Scales
a. Cycloid
b. Ctenoid
7. Adult pigment (lateral)
a. Axial spot beneath pectoral fin
b. Vertical bars
c. Vertical bars festooned
d. Chevrons angled forward
e. Stripes (one or two)
f Inky blotches on lower sides
g. No lateral markings
8. Teeth
a. Conical, widely spaced
b. Flattened, erect, contiguous
c. Flattened, angled backward,
contiguous
Larval characters
9. a. L. jaw with fleshy knob
b. L. jaw without fleshy knob
10. a. Well-marked pigmentation
b. Poorly developed pigmentation
112-123
640
88-92
350
84-88
320
85-96
231
120-128
800
122-128
434
130-155 126-137 145 166
500 467 530 907
+
+
+ +
+ + +
0
0
0
0
0
0
0
0
0
0
+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Fig. 287. Drawings showing changes in pigmentation and body form transformations in Sphyraena barracuda. (A) 17.2 mm SL; collected by
Donald P. de Sylva, 1 mile southwest of the harbor entrance of North Bimini, Bahamas, May 12, 1956. (B) 23.7 mm SL; collected by Donald
P. and Doris D. de Sylva, at beach east of Lemer Marine Laboratory, Bimini, Bahamas, July 7, 1956. (C) 213 mm SL; collected by David K.
Caldwell, Spanish Harbor Key, Monroe County, Florida, June 7, 1956; University of Florida No. 7072. (D) 790 mm SL; collected by Doris D.
de Sylva, north of North Bimim, Bahamas, 25°48'N, 79°17'W, July 18, 1956 (after de Sylva, 1963).
DE SYLVA: SPHYRAENOIDEI
Table 128. Extended.
537
s s
horealis puuditla
S.
sphyraena
S
vthdensts
S S
guachancho putnamtae
S. lello
S. genie
S- afra
S
barracuda
115-130 110-120 120-135 137-140 108-116 108-110 129-131 130-140 120-130 122-140 80-90 1
450 400 1,370 540 470 600 873 1,250 1.150 1.720 1,650 2
+
+
+
+
+
: 1
6
} 7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10
second dorsal fin, and then the pectoral and anal fins (de Sylva,
1963). By 6.6 mm, the second dorsal, artal, and pectoral fins are
fully ossified (Fig. 286). By 1 1.9 mm the first dorsal and pelvic
fins have developed. Middorsal and midventral pigmentation
is well developed at 9 mm, and is useful in differentiating among
larval stages.
Juveniles of most species are unknown, and characters used
to separate adult species would be expected to be the most useful,
especially pigment patterns.
Relationships
Most sphyraenids in museums have been misidentified. The
revision of the family in the Indian Ocean by Williams (1959)
has greatly clarified the identification of several important Indo-
Pacific species whose identification rests largely upon the pattern
of vertical bars or chevrons, festoons (Figs. 288-289), gill raker
characteristics, relative eye size, or upon the relative position
of the first dorsal fin (de Sylva, 1975). The lack of any analysis
of the family based upon osteology, scale morphology (see Bleek-
er, 1854-1857), or internal anatomy precludes an exhaustive
analysis of this family. Electrophoresis and functional enzymic
evolution has related the phylogeny of four eastern Pacific sphy-
raenids to evolutionary temperatures (Graves and Somero, 1982)
and offers much promise for analysis of other fishes. As pre-
viously mentioned, fossil sphyraenids are so incompletely de-
scribed that they shed little light on phyletic affinities.
Because the larvae of only 5 of the 20 species have been
described, almost nothing can be deduced about the phylogeny
of the family based on larval characters.
Sphyraenids were placed by Starks ( 1 900) in the subqrder
Percesoces, together with the Mugilidae and the Atherinidae.
This is based essentially upon their widely separated dorsal fins,
elevated pectoral fins, and the decided branching of the epiotic
crests. Hollister (1937) pointed out that while the mugiloid-
sphyraenoid skeleton superficially resembles that of atherinoids,
there was an important difference in the development of the
hypural plates. Rosen (1964) placed the Atherinoidae in a sep-
arate order, the Atheriniformes. Greenwood et al. (1966) rec-
ognized the Atheriniformes as a superorder, the Atherinomor-
pha, based upon distinctive habits or morphological peculiarities,
and placed in the superorder Acanthopterygii the suborders Mu-
giloidei, Sphyraenoidei, and Polynemoidei.
538
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 288. Variation in lateral pigmentation in various species of Sphyraena. (A) Sphyraena idiasles. 21 cm, Galapagos Islands; (B) Sphyraena
acutipinnis. 27 cm. Hong Kong; (C) Sphyraena novaehollandiae. 43 cm, Kapingamarangi, Caroline Islands; (D) Sphyraena chn'sotaenia, 18 cm,
South Africa; and (E) Sphyraena flavicauda. 33 cm. Strait of Jubal, Red Sea. (All drawn by J. I. Godfrey.)
DE SYLVA. SPHYRAENOIDEI
539
"^^
■-^^^
^'^^::^^~
Fig. 289. Variation in pigmentation in various species of Sphyraena. (A) Sphyraena genie. 29 cm, Makassar, Indonesia; (B) Sphyraena
barracuda, 63 cm, Biscayne Bay, Miami, Rorida; (C) Sphyraena forsleri, 5.9 cm, Indonesia, lateral view; (D) Sphyraena jorsteri, 5.9 cm. Indonesia,
dorsal view; and (E) Sphyraena putnamiae. 45.7 cm, Mahe, Seychelles Islands. (All dravm by J. I. Godfrey.)
540
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
No other close relatives of the Sphyraenidae have been dis-
closed, although it has been postulated that the Australian sea
pike, family Dinolesthidae, is an early offshoot. However, Eraser
(1971) critically compared the internal anatomy of the two fam-
ilies and concluded that their apparent similarity is a result of
convergent evolution.
Larval characters of the Sphyraenidae do not show any ob-
vious similarity to either the Mugilidae or the Polynemidae.
There are only two illustrated accounts of larval Polynemidae
(Aboussouan, 1966d; Kowtal, 1972), neither of which discusses
familial relationships. Superficially, polynemid larvae resemble
the phyletically distant Sciaenidae. Nor do the Mugilidae re-
semble the Sphyraenidae in the larval stages. Undoubtedly there
are similarities in the larval development of the hypural complex
in the Mugilidae and Sphyraenidae, but I am unaware of any
published material on this. The question of whether the poly-
nemids should be grouped within the Mugiloidei and Sphyrae-
noidei is still unresolved.
RosENSTiEL School of Marine and Atmospheric Science,
University of Miami, 4600 Rickenbacker Causeway,
Miami, Florida 33149.
Polynemoidei: Development and Relationships
D. P. DE Sylva
POLYNEMIDAE is the only family of the suborder Poly-
nemoidei, containing 37 species, most of which are Indo-
Pacific. Seven genera have been recognized: Galeoides Giinther,
Filimanus Myers, Pentanemus Giinther, Polynemus Linnaeus,
Polydactylus'LaQtptdt, Polistonemus G'\\\, and Eleutheronemus
Bleeker (see Norman, 1930). This is a shallow-water group
dwelling on sand or mud bottoms, frequently in turbid water.
Most are common in tropical brackish environments, and some
species enter rivers. They are important commercial fishes, es-
pecially in the Indo-Pacific, where some species reach 2 meters.
The threadfins resemble mullets (Mugilidae), but the snout is
pointed and overhangs the large mouth, and the eyes are rather
large (Fig. 290). The feature distinguishing them from their close
relatives, the barracudas and the mullets, is seen in their 4 to 7
pectoral rays which are detached from the rest of the pectoral
fin. Polynemids also differ from mugilids by having a lateral
line, absent in mugilids, which extends onto the caudal fin.
Polynemids are distinguished from sphyraenids by the absence
of fang-like chopping teeth and the rather blunt, terminal mouth
characteristic of the Mugilidae. With the mullets and the bar-
racudas they share the characteristic of 2 widely separated dorsal
fins. The maxillary attachment, shape of the preopercle, length
and number of pectoral filaments, tooth development, and de-
velopment of the lower lip are important taxonomic characters.
Development
Little is known about the eggs and larvae of the Polynemidae
in comparison to the Mugilidae. Eggs have been obtained through
artificial fertilization of Polydactylus se.xfilis in Hawaiian aqua-
culture ponds (Morris and Kanayama, 1964-1969; Lowell, 1971;
Rao, 1977), but illustrations of the egg and larval stages have
not been published. Larval stages of the Indian species Eleu-
theronema tetradactylum from India show developmental stages
from egg to 5.5 mm (Sarojini and Malhotra, 1952; Kowtal,
1972). The small egg, which averages 0.76 mm, has a large oil
globule. In the smallest larva descnbed (3.8 mm), caudal fin
development has started. Some rays appear in the caudal fin at
4.7 mm, and melanophores occur on the maxillary symphysis
and upper side of the pectoral fin bud. A related African species,
Galeoides polydactylus from Senegal, shows little development
of the dorsal fin at 2.7 mm (Aboussouan, 1966d). The head is
relatively large, with a very large eye, and 23 myotomes can be
seen (Fig. 291); they resemble sciaenids. Pigmentation is weak,
in contrast to the Mugilidae, except for some melanophores on
the opercle, anal fin base, and gut. By 4.3 to 4.4 mm, the two
dorsal fins and their rays have formed. At the largest size de-
scribed, 7.6 mm, pigmentation occurs around the opercular se-
ries and posterior trunk, the pectoral filaments are forming, and
the mouth is distinctly inferior. No special larval characters
occur in this group, and development is direct and without any
peculiar metamorphosis.
Relationships
No modem phyletic analysis has been undertaken to delineate
the relationships among the 7 genera. The only revision of the
family is by Gill (1862). The characters which separate them
from one another are the extent of maxillary attachment, shape
of the preopercle, length and number of pectoral filaments, and
development of the teeth and lower lip. Except for the number
of pectoral filaments, those characters at best offer weakly qual-
itative differences useful in identifying species rather than gen-
era.
Early life history stages shed little light on relationships among
members of the Polynemidae. Of the 37 species, larval stages
have been illustrated for only 2 species. Osteological studies on
the axial skeleton have been carried out on 6 species, based
lower pectoral
fin rays
Fig. 290. Major features of the family Polynemidae (from Allen, 1981).
DE SYLVA: POLYNEMOIDEI
541
Table 129. Comparison of Meristic Characters of Mwg;/ and. 4go-
nostomus (Mugilidae), Polydaclylus (Polynemidae), and Sphyraena
(Sphyraenidae) from the Western Atlantic Ocean (Data from
Miller AND JoRGENsoN, 1973).
Fig. 29 1 . Larvae of the polynemoid. Galeoides polydaclylus. (A) 2.75
mm; (B) 3.13 mm; (C) 4.3 mm; (D) 4.4 mm; (F) 7.6 mm. From Abous-
souan(1966cl).
upon adult specimens (Marathe and Bal. 1958). No studies of
the external or internal anatomy have been undertaken on any
polynemid.
The suborder Percesoces was established by Starks (1900) to
show the close relationships among the families Atherinidae.
Mugilidae, and Sphyraenidae. To this group Tate Regan (1929)
added the Polynemidae, based upon the well-developed cranial
crests, the position of the exoccipitals and basioccipitals, the
alisphenoid juncture, the poor development of the parapoph-
yses, and the 24 vertebrae shared with the Sphyraenidae. Based
upon extensive osteological evidence, Gosline ( 1 962) concluded
that the Polynemidae, Sphyraenidae, Atherinidae, and Phallo-
stethoidei are more closely related to one another than to other
fish groups, and placed them in a separate order, Mugiliformes.
He did. however, show that the Polynemidae. Sphyraenidae,
and Mugilidae were more closely related to each other based on
the similar number of vertebrae, the postcleithral strut, and the
possession of nonadhesive eggs, than to the Alhennidae and
Phallostethoidei. The pelvic morphology of the polynemids and
sphyraenids is so primitive as to suggest that these groups could
not have arisen from any advanced percoid groups, and that
they must be derived from a very low level of percoid.
No. of elemenls
Mugd
Agonos-
tomus
dactylus
Sphyraena
Vertebrae
24
25
24
24
Precaudal
Caudal
12
12
12
13
10
14
12
12
First count dorsal fin
Second dorsal fin
5
7-8
5
8
8-9
10-12
6
9
Anal fin
3,8-9
2, 10
3, 12-13
2,9
Total caudal elements
28-29
32-34
41-43
35
Dorsal secondary
Dorsal pnmary
Ventral pnmary
Ventral secondary
7
7
7
7-8
9-10
7
7
9-10
12-13
9
8
12-13
9-10
9
8
9
Gosline (1971) removed the Phallostethoidei from the sub-
order Mugiloidei (the old Mugiliformes), but concluded that the
superfamily Atherinoidae belonged in this suborder together
with the superfamilies Polynemoidae, Sphyraenoidae, and Mu-
giloidae. However, Rosen (1964) had removed the atherinoids
from the percesocine group and had established them as part
of a new, separate order, the Atheriniformes, a practice followed
widely today. Thus, the Polynemoidei, Sphyraenoidei, and Mu-
giloidei have no relatives closer to them than they are to each
other. These are presently recognized as separate suborders within
the order Perciformes.
There is no salient feature in the early life history which relates
the Polynemoidei to other taxa. The development of the eggs
and larvae of Polynemoidei, Mugiloidei. and Sphyraenoidei seem
to follow approximately the same pattern, and all have pelagic
eggs. However, a major departure of the Polynemoidei from the
other two is that one species, Polydaclylus sexfilis. is a protan-
drous hermaphrodite. It matures first as a male at a fork length
of about 20 to 29 cm, and then transforms into the female at
between 30 and 40 cm following a hermaphroditic stage (San-
terreandMay, 1977). As far as is known, none of the Mugiloidei
or the Sphyraenoidei is ever hemaphroditic.
Comparisons of meristic characters offer some noteworthy
data (Table 129). The vertebral count of the Polynemoidei is
10 + 14 = 24; the other two suborders have a count of 12 -I-
12 = 24. The number of dorsal and anal elements of the Mu-
giloidei and Sphyraenoidei resemble each other more closely
than they do the Polynemoidei. The vertebral formula, as well
as the number of dorsal and anal elements, are more closely
related to the Gerreidae. In fact, the habits of the Polynemoidei
closely resemble those of the Gerreidae. To my knowledge, there
is nothing published on the eariy life history of the Gerreidae
which might disclose any similarities to the Polynemidae.
The Polynemoidei (i.e.. Polydaclylus) have a higher number
of first dorsal, second dorsal, anal, and caudal elements than
the other groups (Table 129). However, a companson on a
worldwide basis is required before such an analysis can reveal
phyletic relationships.
RosENSTiEL School of Marine and Atmospheric Science,
University of Miami, 4600 Rickenbacker Causeway,
Miami, Florida 33149.
Labroidei: Development and Relationships
W. J. Richards and J. M. Leis
THE most recent concept of this group concludes that the
Pomacentridae, Cichhdae, Embiotocidae, and Labridae
comprise a monophyletic assemblage (Kaufman and Liem. 1982).
Kaufman and Liem (1982) include the Odacidae and Scaridae
in the expanded family Labridae. For present purposes, we em-
ploy the traditional view of three separate families. Pomacen-
tridae is a large primarily marine family of about 23 genera and
230 species found in the tropics and warm temperate waters of
the world's oceans (Allen, 1 975a). Cichhdae is a fresh and brack-
ish water family found in the Americas, Africa including Mad-
agascar, coastal western Mediterranean, and the coastal areas
of India. It is a large family comprised of about 85 genera and
perhaps 700 species making it the second largest perciform fam-
ily (Stiassny, 1981). Embiotocidae is found only in the North
Pacific with 2 species around Japan and Korea. 18 off the west
coast of the LI.S., and 1 confined to freshwater of California
(Tarp, 1 952). Labridae is a tropical and warm temperate marine
family of about 58 genera and about 400 species (Russell, 1 980).
Odacidae is a temperate marine group of 4 genera and 1 2 species
confined to New Zealand and southern Australia (M. F. Gomon
and J. R. Paxton, pers. comm.). Scaridae is a tropical marine
family of about 12 genera and 70 species (Schultz, 1958). Table
1 30 summarizes meristic characters of labroid fishes.
Development
The family Embiotocidae is a small family of viviparous species
that has several unusual morphological specializations during
development as reviewed by Wourms (1981). During gestation,
the vertical fins hypertrophy and develop spatulate extensions,
and the alimentary canal hypertrophies, especially the hind gut.
All these specializations appear due to viviparity and are not
treated further here.
Cichhdae, so far as known, all undertake elaborate parental
care (Breder and Rosen, 1966). The eggs are slightly elliptical
or irregularly shaped. The eggs are also adhesive except for those
which are orally incubated. There is a vast literature on repro-
ductive behavior most of which describes spawning behavior
and parental care, but there is little descriptive information on
larvae since many species transform directly from yolk sac to
juvenile (Balon, 1981b; Noakes and Balon, 1982). Balon (1959)
described the young of Cichlasoma cyanoguttalum. The larvae
of laboratory reared Cichlasoma octofascialum are depicted in
Fig. 292. The larvae of Pterophy/him have an adhesive disk on
the head for attachment to substrate and several stages are de-
picted in photographs in Innes (1956). The Cichlasoma larvae
(Fig. 292) have unusual structures on the head though they were
not observed to be used as holdfast organs (A. W. Kendall,
pers. comm.). Larvae of Symphysodon cling to the mucus of the
parent and actually derive nourishment from it (Breder and
Rosen, 1966). Balon (1977) thoroughly describes the develop-
ment of Labeotropheus, a mouth brooder which has direct de-
velopment.
Pomacentridae have demersal eggs with an adhesive pedestal;
the male guards and incubates them. Few species have been
studied from an early life history perspective (Table 131). Most
have pelagic larvae, but at least one species (Acanthochromis
polyacantha) broods and protects the young in a manner similar
to cichlids (Robertson, 1973). Larval development is direct with
few larval specializations and no specialized stages between lar-
vae and juveniles. The sequence of fin formation is variable.
All fins may be formed as early as 3 mm, but depending on
species, settlement may not occur until 1 8 mm. The gut is coiled
at hatching. The larvae arc very similar to percoids and may
be easily confused with numerous families (Leis and Rennis,
1983). In general they have a short, coiled, triangular gut, an
inconspicuous gas bladder which is covered by melanophores,
and weak preopercular spination (Fig. 293).
Some early life history information is available on about one-
half of the labrid genera (Table 1 30). The vast majority of labrids
spawn small (0.5-1.1 mm) pelagic eggs, but three northeast
Atlantic genera have adhesive, demersal eggs with parental care
(Table 131). Demersal labrid eggs are small (< 1 mm) and ad-
hesive, but do not have an adhesive pedestal. Labrid eggs usually
have a smooth chorion and a single oil globule. Newly hatched
larvae have the yolk sac protruding anteriorly in front of the
head with the oil globule (if present) at the anteriormost postion.
The larvae are generally elongate and laterally compressed with
a deep caudal peduncle, but some species are deep-bodied (Fig.
294). The gut is rugose and is initially straight; coiling may be
delayed until after flexion in some species. The head is com-
pressed and almost always lacks spines. Scales do not form prior
to settlement. In tropical forms the eye may be round, ovoid.
Table 130. Some Meristic CHAEtACTERS of Labroid Fishes. N is the approximate number of recent species largely after Nelson (1976). Other
data from Gunther, 1862; Boulenger, 1915; Tarp. 1952; Miller and Jorgensen, 1973; Russell, 1980; Leis and Rennis, 1983; Sanchez, 1981; and
J. R. Paxton, pers. comm.
N
D
A
F
'
Pj
Venebrae
Cichlidae
700
IX-XXV, 3
-31
III
-XIII, 6-28
_
_
1,5
24-39
Embiotocidae
23
VII-
XVIII,
9-28
III
-IV, 13-35
17-
-29
1,5
31-42
Labridae
400
VIII
-XX, 5
-15
III
-V, 6-14
11-
-21
1,5
23-40
Odacidae
11
XIV
-XXVI
9-23
II-
111,8-14
11-
-18
0, 0 or I, 4
31-54
Pomacentridae
230
VIII
-XVII,
10-18
II,
10-18
14-
-22
1,5
26
Scaridae
70
IX,
10
III
8-9
13-
-17
1,5
26
542
RICHARDS AND LEIS: LABROIDEI
543
Fig. 292. Larva of (upper) Cichlasoma oclofascialum. 5.4 mm SL, laboratory .reared, 4 days after hatching, drawn by B. Vinter and (lower)
5.9 mm SL, laboratory reared, 10 days after hatching, drawn by B. Vinter.
squarish, or nairow and have choroid tissue associated with it.
Larvae of temperate species tend to have heavy melanistic pig-
ment while tropical forms have few melanophores although
erythrophores may be abundant. Meristic characters are very
useful for identifying these larvae. Development is direct, with
only the non-round eyes (some with choroid tissue) and elongate
fin rays of some species, and perhaps the reduced melanistic
pigment of tropical taxa representing lai-val specializations. Most
Fig. 293. Lar\'a of Microspathodon chrysurus 3.7 mm SL from specimen reared in the laboratory. From MS by Potthoffet al., drawn by J.
C. Javech.
Table 131. Labroid Taxa for Which Information is Available on Ego and Larval Stages. References dealing with spawning which do
not describe eggs or larva are omitted. YS— yolk-sac stage; pre^preflexion stage; flex — flexion stage; post — postflexion stage; D— demersal, P—
pelagic.
Family/genus
Number of
species
Larvae— developmental stage
Egg type
Pre
Flex
References
Pomacentridae
Abudefduf
Acanthochromis
Amphipnon
Chromis
Scaridae
Calotomus
NicholsinaC!)
Scarus
Spansoma
Unidentified
Odacidae
Neoodax
Odax
D
D
D
D
Several
P-round
P-spindle
P-round
P-spindle
X
Parental care of larvae
X X X X
X
Microspathodon
1
D
X
X
X
X
Pomacenlrus
1
D
X
X
X
Stegastes
1
D
X
Unidentified
Several
X
X
X
X
Labridae
Bodiamts
1
X
Cenlrolabrus
1
D
X
Cheilinus
1
X
Choerodon
1
X
Cirrhilabrns
1
P
X
Coris
2
P
X
X
X
X
Clenolabrus
1
P
X
X
X
X
Halichoeres
4
P
X
X
X
Imistius
1
X
Labroides
1
P
X
Labrus
2
D
X
X
X
X
Lachnolaimus
1
P
X
X
X
X
Novaculichthys
1
X
Oxyjulis
1
P
Pseudocheilinus
1
X
Pseudolabrus
3
P
X
X
X
Pleragogiis
1
P
X
Semicossyphus
1
P
X
StethojuUs
2
P
X
X
X
Symphodus
6
D
X
X
Tautoga
1
P
X
X
X
Taulogolabrus
1
P
X
X
X
Thalassoma
4
P
X
X
X
Xyrichthys
3
P
X
X
Unidentified
Several
P
X
X
X
X
X
X
X
X
X
X
Shaw. 1955; Mito, 1966; Miller
et al.. 1979; Re, 1980; Leis
and Rennis, 1983
Robertson, 1973
Delsman, 1930c; Allen, 1972;
Vatanachi, 1972; Leis and
Rennis, 1983
Page, 1918; Padoa. 1956d; Fujita,
1957a; Turner and Ebert, 1962;
Ahlstrom, 1965; Leis and
Rennis, 1983
Potthofl'et al., MS
Leis and Rennis, 1983
Miller etal., 1979
Nellen, 1973b; Leis and Rennis,
1983
Richards, 1984
Sparta, 1956a; Russell, 1976
Leis and Rennis, 1983
Leis and Rennis, 1983
Suzuki et al., 1981
Page, 1918; Sparta, 1956a;
Fourmanoir, 1976
Russell, 1976
Mito, 1962b; Pourmanoir, 1976;
Richards, 1984
Masuda and Tanaka, 1962
Suzuki etal., 1981
Sparta, 1956a; Russell, 1976
Colin, 1982; Kelley, pers. comm.
Vatanachi, 1972
Bolin, 1930; Orton, 1953a
Fourmanoir, 1976
Mito, 1962b; Robertson, 1975a;
Crossland, 1981
Mito, 1962b
Orton, 1953a
Mito, 1962b; Nellen, 1973b
Sparta, 1956a; Russell, 1976
Kuntzand Radclifle, 1917
Kuntzand Radclifle, 1917
Kubo, 1939; Sparta, 1956a; Leis,
1983; Richards, 1984
Sparta, 1956a; Leis and Rennis,
1983; Richards, 1984
Kamiya, 1925; Mito, 1962b; Dekhnik
et al., 1966; Pourmanoir,
1976; Miller etal., 1979;
Crossland, 1982; Leis and
Rennis, 1983
Kamiya, 1925; Leis and Rennis,
1983
Regan, 1916; Aboussouan, 1969
Winn and Bardach, 1960; Mito,
1962b
Sparta, 1956a; Winn and Bardach,
1 960; Randall and Randall,
1963
Watson and Leis, 1974; Leis
and Rennis, 1983; Richards,
1984
Regan, 1916
Robertson, 1975a; Crossland,
1982 (as unidentified larva 1)
RICHARDS AND LEIS: LABROIDEI
545
->•>
Fig. 294. Labrid larvae from top to bottom: Lachnolaimus maximus. 5.0 mm SL, from specimen reared in the laboratory (from MS by Kelley)
(drawn by J. C. Javech); Thalassoma bifasciatum 8.2 mm SL collected on R/V OREGON II cruise 7239. station 10, 18°00'N latitude, 059°59'W
longitude, July 14, 1972 (drawn by B. Washington); Xynchlhys sp. (deep body form), 5.0 mm SL, collected on Ry'V OREGON II cruise 7239,
station 149, 23°29'N latitude, 079° 1 3' W longitude. August 7. 1972. [Note narrow eyes. Freshly caught specimens havered pigment (erythrophores)
on the head, trunk, and tail] (drawn by B. Washington); and Xynchthys sp. (narrow body form), 10.5 mm SL, collected on R/V OREGON II cruise
7239, station 149, 23°29'N latitude, 079°13'W longitude, August 7. 1972. Note narrow eyes. Freshly caught specimens have red pigment
(erythrophores) on the head, trunk, and tail (drawn by B. Washington).
546
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
Fig. 295. Larva (upper) of an unidentified scarid, 9.3 mm SL, collected on RA' OREGON II cruise 7239, station 54, 18°58'N latitude, 080°09'W
longitude, July 30, 1972; [meristic characters for Atlantic scarids and the labrid Doralonotus megalepis are identical] (drawn by B. Washington):
and larva (lower) of Oda.x pul/us, 12.2 mm SL, from New Zealand (drawn by J. C. Javech).
labrid lai^ae settle out at less than 1 5 mm, but some may remain
pelagic until 25 mm.
Scarids spawn pelagic eggs: the subfamily Scarinae appears
to spawn spindle-shaped eggs, and the subfamily Sparisomatin-
ae to spawn spherical eggs (Table 131). Morphologically, scarid
larvae are similar to many labrids: they are elongate and com-
pressed; have an initially straight, rugose gut that later coils;
lack head spines; have squarish to narrow eyes; and usually
develop choroid tissue (Fig. 295). Scarid larvae differ most strik-
ingly from labrid larvae in melanistic pigment. Scarid larvae
consistently have melanophores over the posterior gut and have
a ventral series of melanophores on the tail. Melanophores in
the cardiac region and dorsally on the caudal peduncle are com-
monly found in scarids. Melanophores in these regions are either
absent or limited to one or two melanophores in tropical labrids.
The ventral series of melanophores on the tail of some scarid
larvae resembles a set of developing photophores (a histological
study is warranted). This ventral pigment plus the narrow eyes
and choroid tissue (particularly of sparisomatines) give some
scarid larvae a gonostomatid or myctophid appearance, result-
ing in some identification problems. Scarines seem to settle out
at 10 mm or less, while Calatomus (a sparisomatine) may re-
main pelagic until 15 mm.
Little is known of the early life history of odacids, but they
spawn pelagic eggs (Table 131), and their larvae are generally
similar to elongate labrids with high numbers of myomeres
(Table 131). Only three larvae of two species have been de-
scribed, so it is difficult to generalize, but these are elongate.
compressed, have unlooped guts, no head spines, and round
eyes. One species has very elongate, early-forming, anterior spines
in the dorsal fin, and a pigment pattern of blotches along the
body margins (Fig. 295). The other species is unpigmented and
lacks elongate fin elements.
Relationships
Kaufman and Liem (1982) include in the Labroidei the Po-
macentridae, Cichlidae, Embiotocidae, Labridae, Odacidae, and
Scaridae and further include the Odacidae and Scaridae in the
Labridae. They consider the Pomacentridae to be the primitive
sister group of all the other labroids, the cichlids the primitive
sister group of embiotocids and labrids, and embiotocids the
primitive sister group of the labrids.
Labroids are characterized by (1) united or fused fifth cera-
tobranchials resulting in the formation of one functional unit,
(2) a true diarthrosis between upper pharyngeal jaws and the
basicranium without an intervening part of the transversus dor-
salis anterior muscle, and (3) the presence of an undivided
sphincter oesophagi muscle forming a continuous sheet (Kauf-
man and Liem, 1982).
Kaufman and Liem's (1982) arrangement and composition
of the Labroidei receives only limited support from ELH char-
acters. The monophyly of the Labroidei cannot be established
from early life history characters. Pomacentrid and cichlid lar-
vae are morphologically and developmentally nearly indistin-
guishable from many percoid larvae (e.g., mullids, gerreids,
sparids), while the labrids, scarids, and odacids are quite dif-
RICHARDS AND LEIS: LABROIDEI
547
ferent. A cursory study indicates larvae of these latter families
share at least four derived characters: almost total lack of head
spination; a long, rugose, straight gut which loops relatively late
in development; compressed, elongate body; and a reduction in
principal caudal ray number from the typical percoid comple-
ment of 9 + 8. The "percoid" larval type of the pomacentrids
and cichlids might be a primitive character state, but there are
no derived characters which unite their larvae with the labrid
type of larvae. At least gut development and head spination of
the labrids are shared with the pseudochromids, which are gen-
erally very similar to some labrid larvae which settle at small
sizes (Leis and Rennis, 1983). This may be the result of con-
vergence, but a labrid/pseudochromid relationship should be
investigated as an alternative to Kaufman and Liem's (1982)
proposed phylogeny.
If Kaufman and Liem (1982) are correct in proposing the
pomacentrids as the primitive sister group of the other labroids,
then either parental care of hatched young evolved indepen-
dently in the pomacentrid Acanihochromis and the cichlids (vi-
viparity in embiotocids might be a derivation of parental care
of eggs and hatched larvae, but this remains to be shown), or
was present in a pre-pomacentrid common ancestor and was
secondarily lost in all labroids but the cichlids and Acaniho-
chromis. Similarly, either demersal eggs and parental care of
them evolved independently in some labrids, pomacentrids, and
cichlids. or were present in a pre-pomacentrid common ancestor
and secondanly lost in most labrids and all scarids and odacids.
Therefore, neither demersal eggs nor parental care of hatched
young offer much support to Kaufman and Liem's (1982) phy-
logeny.
ELH characters may be useful in studying the intrafamilial
relationships of labroid fishes. Larval labrids are very diverse
in development and morphology, and this may prove useful in
elucidating labrid interrelationships. Within the labrids, de-
mersal eggs and parental care of eggs are unique to some mem-
bers of the tribe Labrini. Egg shape and larval morphology sup-
port the subfamilial divisions within the Scaridae. Too little is
known of pomacentrid and cichlid development to say if ELH
characters might be useful in elucidating intrafamilial relation-
ships.
In conclusion, ELH characters support Kaufman and Liem's
(1982) labroid phylogeny only in the close relationship of the
labrids, scarids, and odacids. In spite of the similarities uniting
the three families, there are enough differences between their
known larvae to lead us to suggest the labrids, scarids, and
odacids should not be combined into one family at this time.
M. F. Gomon (pers. comm.) argues that the alternative to com-
bining the three families into one is splitting the group into as
many as five smaller families. While we do not advocate this
course, the great larval diversity found within the group could
provide evidence supporting this alternative. However. ELH
information for more genera of labrids, scarids, and odacids
must be gathered before firm statements can be made.
(W.J.R.) National Marine Fisheries Service, Southeast
Fisheries Center, 75 Virginia Beach Drive, Miami,
Florida 33149; (J.M.L.) The Australian Museum, 6-8
College Street, Sydney 2000, Australia.
Acanthuroidei: Development and Relationships
J. M. Leis and W. J. Richards
THE acanthuroid fishes are marine, tropical and, except for
the pelagic Luvaridae, are associated with coral reefs. The
suborder consists of about 1 10 species distributed among four
families: Acanthuridae (Randall, 1955a), Luvaridae (Roule,
1924; Tyler, Nakamura and Collette, MS in prep.). Siganidae
(Woodland, 1983), and Zanclidae (we follow Randall, 1981, and
consider the Zanclidae distinct from the Acanthuridae). Ap-
parently, all species have a specialized pelagic stage between
larvae and juveniles, often referred to as an acronurus larval
stage (we prefer to restrict this term to its original usage in the
Acanthuridae). This specialized pelagic stage has provided the
basis for the description for many supposedly new species and
genera, and has been used as evidence for uniting the group
(e.g., Lauder and Liem, 1 983). The siganids are usually consid-
ered the most generalized (=primitive) family of the suborder,
and the zanclids are considered closely related to if not included
in the acanthurids (Tyler, 1 970). Luvanis has recently been shown
to be closely related to the acanthurids (Tyler, Nakamura and
Collette, MS in prep.). The chaetodontids have been suggested
as the percoid group from which acanthuroids were derived
Table 132. Meristic Characters of Acanthuroid Fishes. N is the number of recent species, principally after Nelson, 1976. Note that in the
Luvandae there is a progressive loss offin rays from the larval stage (adult counts in parentheses). Maximum larval counts are followed by adult
counts in parentheses. (Data from Randall. 1955b. c: Smith. 1966a; Weber and de Beaufort. 1936; Gregory and Conrad. 1943; and Leis and
Rennis. 1983).
N
D
A
P.
P;
c
Venebrae
Acanthundae
85
IV-IX. 19-33
II-IV. 18-32
14-19
I. 3-1, 5
16
22-23
Siganidae
23
XII-XIV. 9-11
VII. 9-10
14-21
1,3,1
17
23
Zanclidae
1
VII. 38-42
111.31-35
18-19
1.5
16
22
Luvaridae
1
11.24(12-13)
18(13-14)
17-20
1.4(0)
16
23
548
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 133. Acanthuroid Taxa for Which Information is Available on Ego and Larval Stages. YS— yolk-sac; pre— preflexion; flex-
flexion stage; post — postflexion stage; D— demersal; P— pelagic.
Number of
species
Larvae— developmental stage
Egg type
Pre
Flex
Acanthuridae
Acanihurus
Clenochaetus
Naso
2
2
Zebrasoma
Unidentified
2
Several
Luvaridae
Luvarus
1
Siganidae
Siganus
6
Zanclidae
Zanchis
X
X
X
X
X
X
Lutken, 1880; Breder. 1927; Whitley
andColefax, 1938; Randall, 1956,
1961; Aboussouan, 1965; Burgess,
1965
Randall, 1955c
Fourmanoir, 1976; Leis and Rennis,
1983
Randall, 1955b; Aboussouan, 1966a
Dekhnik et al., 1966 (Fig. 37-1, misiden-
tified as Balistidae); Randall, 1955c;
Nellen, 1973b; Watson and Uis, 1974
Roule, 1924; Roule and Angel, 1930;
Blache, 1964
Fujita and Ueno, 1954; Uchida et al.,
1958; Mito, 1966; Popper etal., 1973;
May et al., 1974; von Westemhagen
and Rosenthal, 1975, 1976; Bryan and
Madrisau, 1977; Leis and Rennis,
1983
Strasburg, 1962
(Tyler, 1970). Table 132 summarizes meristic characters of the
suborder, and Table 133 reviews current state of knowledge of
its early life history.
Development
Siganids have small (< 1 mm) demersal eggs with oil droplets
(Table 133). No parental care has been recorded. Larvae hatch
in a poorly developed condition. Moderately long preopercular
spines and serrate ridges form on the head, and the first fin
elements to form— the pelvic spine and second dorsal spine—
form very early (Fig. 296). The body does not become very
deep, and although the pelvic and dorsal spines are elongate
and serrate, they do not exceed three times the eye diameter.
No scales form prior to settlement, but the pelagic stage may
grow to 30 mm and is very silvery in life, particularly over the
gut. Early larvae, in particular, are very percoid in appearance.
Figment in preflexion larvae is limited to dorsal and pelvic fin
membranes, gut, and a ventral series on the tail. Older larvae
are more heavily pigmented.
Acanthurids have small (< I mm) pelagic eggs with a single
oil droplet (Table 133). Larvae hatch in a poorly developed
condition, but very soon develop serrate ridges on the head (but
no elongate preopercular spines form). The first fin elements to
form (the pelvic spine and second dorsal spine) do so very early,
and these are quickly followed by the second anal spine (Fig.
296). These fin spines are serrate, and at least one exceeds three
times the diameter of the eye. The head and trunk become
remarkably deepened. This is accentuated by the elongate pel-
vic, dorsal, and anal spines at the opposite edges of the deepest
point of the body: the body becomes distinctly kite-shaped.
Small, triangular scales arrayed in vertical rows begin to form
shortly after flexion. The pelagic stage may reach 60 mm and
is very silvery in life around the gut. Preflexion larvae are lightly
pigmented in specific patterns. Late larval stages may acquire
aspects of the juvenile pigment pattern. The caudal peduncle
armature forms late in the larval period. In Naso the spines form
from existing scales (i.e., they pass through an unspecialized
scale stage). In Acanihurus it forms directly without the unspe-
cialized scale stage.
Nothing is known of luvarid or zanclid eggs or preflexion
larvae (Table 133). Luvarus larvae apparently have early-form-
ing pelvic and anterior dorsal fin spines. They also have early-
forming scales, serrations on the head, but lack elongate pre-
opercular spines (Fig. 297). The dorsal and pelvic spines of
Luvarus are more than three times the diameter of the eye.
Luvarus larvae are deep-bodied, but not as kite-shaped as acan-
thurids, and have a more square-shaped head. With growth, the
spines of the fins, and many of the soft rays are lost, and the
body becomes more fusiform. Late Zanclus larvae are very
similar to acanthurid larvae (Fig. 297) and are scaled similarly.
Fig. 296. Urvae of (upper) Acanihurus sp., 6.0 mm SL, OREGON II cr. 7343, sta. 87, Caribbean Sea, 16''54'N. 062°03'W, February 16.
1973, drawn by J. C. Javech; (middle) Naso unicornis. 5.9 mm SL, PROVIDENCE II, st. T-429, Indian Ocean, 09°27'S, 050°2rE, December
18, 1974, drawn by J, C. lavech; and (lower) Siganus fuscescens, 5.4 mm TL, modified after Uchida et al. (1958).
LEIS AND RICHARDS: ACANTHUROIDEI
549
550
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 297. Larvae of (upper) Luvarus imperiatis, 6.8 mm TL, modified after Fahay ( 1 983); and (lower) Zanclus canescem. 16 mm SL. modified
after Strasburg (1962).
but only the third dorsal spine is elongate (the unknown smaller
larvae may have elongate spines in other fins).
Relationships
Acanthuroids share the following, probably derived charac-
ters (we assume acanthuroids have a percoid ancestry): long
pelagic period; early-forming, elongate dorsal and pelvic spines;
serrate fin spines; moderately to very deep, compressed body;
serrate ridges on the head; silvery gut; 22-23 vertebrae; and 16-
1 7 principal caudal rays. This is strong evidence for the
monophyly of the group.
Tyler (1970) notes that acanthuroids have been considered
as chaetodontid derivatives. We find no support for this view
among ELH characters. Chaetodontids and pomacanthids do
have large, specialized pelagic stages, but these differ greatly
from acanthuroids (Leis and Rennis, 1983) and resemble ca-
LEIS AND RICHARDS: ACANTHUROIDEI
551
rangids at early stages. Leiognathid larvae are similar to siganid
larvae in many respects (head spination, fin spine development,
silvery gut— see G. D. Johnson, this volume and Leis and Gold-
man, 1 983), and we suggest the leiognathids should be evaluated
as a potential primitive sister group of the acanthuroids. There
is little evidence from ELH characters to support the notion
that the acanthuroid fishes are the primitive sister group of the
tetraodontiform fishes (Leis, this volume).
Intra-ordinal relationships of acanthuroid fishes as suggested
by ELH characters fully support those based on adult characters.
The siganids are distinguished from the other acanthuroids by
the following derived characters: demersal egg, two spines in
pelvic fin, and seven spines in the anal fin. Larvae of acanthurids,
luvarids, and zanclids have the following derived characters: no
elongate preopercular spines; kite-shaped body; elongate snout;
extremely elongate dorsal and pelvic spines; early-forming spe-
cialized scales; and reduced number of dorsal fin spines. Thus
the siganids appear to be the primitive sister group of the other
acanthuroids. Interrelationships of the acanthurids, zanclids,
and luvarids cannot be clarified given the current knowledge of
zanclid and luvarid ELH characters. Larval zanclids have an
extremely elongate dorsal spine and a retrose preorbital spine.
Acanthurids have caudal peduncle armature, and luvarids have
ontogenetic reduction in fin elements, no anal spines, and a very
squared head. None of these specializations are shared by any
two of the families, so they shed no light on interrelationships.
(J.M.L.) The Australian Museum, 6-8 College Street,
Sydney 2000, Australia; (W.J.R.) National Marine
Fisheries Service, Southeast Fisheries Center, 75 Vir-
ginia Beach Drive, Miami, Florida 33149.
Blennioidei: Introduction
R. H. Rosenblatt
THE modem concept of the perciform suborder Blennioidei
dates from the paper of Regan (1912b), who defined and
delimited the group as "Percomorphous Teleosts with the pelvic
fins jugular or mental, each of a spine and four soft rays or still
further reduced, with the dorsal and anal rays typically corre-
sponding in number to the vertebrae, each basal bone attached
to its own neural or haemal spine (rays more numerous in Ophi-
diiformes) with well developed wings of the parasphenoid as-
cending in front of the prootics, and with all or most of the ribs
inserted on strong parapophyses."
As Regan himself indicated this definition encompasses a
heterogeneous group, and his series "Ophidiiformes" has now
been removed from the Perciformes. Subsequent to Regan sev-
eral widely differing classifications have been proposed, with
groups often being added or removed without comment. Jordan
(1923) proposed the most radical arrangement. He placed in the
order Jugulares almost all spiny rayed fishes with advanced
pelvic fins. Jordan's Jugulares was divided by him into 1 2 series,
comparable to suborders, and no less than 62 families. Jordan,
in his magisterial fashion, provided an outline classification,
without substantiation by characters.
Berg in his 1940 classification rationalized the classification
of the Blennioidei. He restricted the suborder mainly to Regan's
series "Blenniformes" and "Cliniformes," and redistributed the
remainder of the Jugulares, either to the Percoidei or to the
suborders Ophidioidei (equivalent to Regan's series Ophidi-
iformes), Ammodytoidei, or Callionymoidei. Some indication
of relationships is perhaps implicit in Berg's placement of por-
tions of Jugulares auctorum immediately preceding the Blen-
nioidei.
Although a number of works on various blennioid groups
have appeared (see particularly Hubbs, 1952; Makushok, 1958,
and the papers of V. Springer) the only subsequent attempt to
characterize and deal with the group as a whole is that of Gosline
(1968). The classification given by Nelson (1976) differs from
that of Gosline as well as the outline classification of Greenwood
et al. (1966). The discussion of larval forms given here mostly
accords with Nelson's Blennioidei as a convenience, regardless
of the eventual disposition of the taxa. The only major departure
from the arrangement of Nelson is that the family Zoarcidae is
treated here, although Nelson included it in the Gadiformes (see
Anderson, this volume).
The reasons for the varying treatment of these fishes are not
difficult to find. The unraveling of phyletic lines within the
Perciformes is made difficult by the sheer number of species
and genera. One is faced with the choice of mining a narrow
vein for nuggets of knowledge which lie isolated, or engaging in
a strip mining operation which reveals broad patterns at the
expense of ignoring contradictory details. In other terms, in-
sufficient knowledge of morphological variation within the Per-
ciformes precludes at this time either identification of unequiv-
ocal synapomorphies or the determination of polarity of a number
of characters within almost any presumed lineage.
A number of features taken to characterize, if not to define,
the Blennioidei may be the product of convergent or parallel
evolution, correlated with the assumption of benthic life.
As pointed out by Gosline ( 1 968) the blennioids, as compared
with percoids, have less deep bodies, with a short trunk and a
relatively attenuated caudal region. The dorsal and anal are long
and low, terminating near the caudal, and the pectoral and usu-
ally the caudal fins are rounded. There is an exact correspon-
dence in number between dorsal and posterior anal soft-rays
and vertebrae supporting them. The pelvic fins are inserted in
advance of the pectoral fins, and the number of rays is generally
reduced; the spine often rudimentary or splint-like, and the soft-
rays three or fewer.
The deep, relatively compact body of a generalized perciform
is that of a fish which hovers, probably near the substrate, but
which makes rapid bursts either in feeding or predator avoid-
ance, or both. The body shape is adapted for slow swimming.
552
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
alternating with bursts of acceleration. There is a general but
far from universal trend for bottom dwelling fishes to become
elongate; the eel-like body is widely distributed taxonomically.
Bottom-living fishes often use crevices for shelter and forage in
interstices, and may burrow. Elongation of the body accom-
panied by an increase in the number of vertebrae produces the
flexibility necessary for these activities. The elongate body form
requires either anguilliform swimming or undulation of the me-
dian fins. In either case the role of the caudal fin is reduced.
The pectorals are used in short darts or lunges, and their fan
shape is associated with accelerating a large amount of water
per thrust. This function is important even in relatively elongate
forms in accelerating the head in feeding strikes, and pectorals
are reduced or lost in only a few lineages.
The pelvics of bottom living forms no longer have a hydro-
dynamic function as brakes or rudders. Instead they may func-
tion as props which hold the head ofl^ the bottom (as in the
Cottidae and Gobiidae as well). A reduction in the number of
rays is also seen in the Cottidae.
That morphological features are functional does not mean
that their joint possession cannot be taken to demonstrate com-
mon ancestry. However, it does indicate caution. The only one
of Gosline's characters for the Blennioidei that is not clearly
functional is the 1:1 relationship of median fin rays and ver-
tebrae. However, the reduction in the number of fin rays per
segment to one is the culmination of a functional trend begun
in the Paleozoic, and we cannot yet be sure that it happened
but once.
Although Gosline regarded his classification as owing more
to that of Jordan than Regan, his mam characters of pelvic
position and median fin ray arrangement are exactly those given
by Regan in his diagnosis. Gosline's concept of the Blennioidei
and its superfamilies, although not completely accepted (see
Nelson, 1976), has not been superseded, except that his Con-
grogadoidae is no longer included; the Congrogadidae is now
placed in the Percoidea (Winterbottom, 1982) and the Pero-
nedysidae has been synonymized with the Clinidae (George and
Springer, 1980).
According to Gosline the Blennioidei (without the Congro-
gadoidae) may be divided into four superfamilies. The first of
these, the Notothenioidae, is clearly the most heterogeneous. In
addition to the Antarctic and sub-Antarctic families (Bovich-
thyidae, Nototheniidae, Harpagiferidae, Bathydraconidae and
Channichthyidae) usually placed in this group (Berg, 1940), the
tropical Mugiloididae (=Parapercidae) Trichonotidae and Chei-
marrhichthyidae were included although they do not share with
them the specialized features of a smgle nostril and a loss of
one pectoral actinost. There appears to be no reason to regard
the two groups of families as closely related.
The Trachinoidae was said to be comprised of the Trachin-
idae, Leptoscopidae, Uranoscopidae and Dactyloscopidae. All
are adapted for lying buried in the substrate, and it is likely that
their structural similarities are related to this habit. The Dac-
tyloscopidae has recently been placed in the Blennioidae (George
and Springer, 1982).
The superfamily Blennioidae was regarded as composed by
the families Tripterygiidae, Clinidae, Chaenopsidae, and Blen-
niidae. Subsequently the subfamily Labrisominae of the family
Clinidae was raised to family status and the Dactyloscopidae
transferred from the Trachinoidei (George and Springer, 1980).
Within the Blennioidei, the superfamily may be characterized
by the combination of two nostrils on each side, pelvic soft-
rays four or fewer, prootic excluded from orbital rim (that is,
ascending wing of paraspheroid meets frontal), and basisphe-
noid present.
The remaining superfamily, the Zoarceoidae, was regarded as
composed of 1 1 families, some poorly understood. Anderson
(1983, this volume) recognized 8 families in the group: Bathy-
masteridae, Stichaeidae, Pholididae, Anarhichadidae, Ptil-
ichthyidae, Zaproridae, Scytalinidae and Zoarcidae. Although
composed of forms differing greatly in morphology, the super-
family may be diagnosed as blennioids with a single nostril on
either side of the head, prootic excluded from rim of orbit, and
basisphenoid absent. There is no merit in the removal of the
Zoarcidae to the Gadiformes (see also Anderson, this volume).
It should be clear from the foregoing that no satisfactory def-
inition of the Blennioidei has as yet been framed. Perhaps lines
of relationships would best be recognized by restricting the Blen-
nioidei to the Blennioidae and Zoarceoidae of Gosline, and
returning his other two superfamilies to the Percoidei. It appears
that ontogeny and larval characters have as yet little to con-
tribute to questions of suprafamilial and subordinal relation-
ships between and among these fishes.
Perhaps it is fitting to end with a quote from Jordan (1923),
addressing issues such as this: "I may repeat a warning as old
as science itself: that we must not expect a degree of accuracy
which the subject in question does not permit."
ScRipps Institution of Oceanography, University of Cal-
ifornia, San Diego, La Jolla, California 92093.
Schindlerioidei: Development and Relationships
W. Watson, E. G. Stevens and A. C. Matarese
THIS suborder contains a single paedomorphic family com-
posed of two species of the genus Schindlena. Both species
inhabit neritic surface waters of the subtropical and tropical
Indian and Pacific Oceans (Bruun, 1940; Schindler, 1932; R. J.
Lavenberg, pers. comm.). Their early life histories are known
from the work of Watson and Leis (1974), Miller et al. (1979),
and Ozawa and Matsui (1979). Classification of Schindlerioidei
is speculative, and its placement here by Nelson (1976) follows
Gosline (1971), who tentatively considered this taxon a percoid
derivative, possibly related to Ammodytoidei.
WATSON ET AL.: SCHINDLERIOIDEI
553
B
.:^:^Za:Z^^^^
Fig. 298, Lateral views of: (A) Schmdlena pietschmanni larva, 2.7 mm; (B) 5. pietschmanni larva. 3,5 mm (redrawn from Miller et al.. 1979);
(C) 5. pietschmanni larva. 4.7 mm (from Miller et al., 1979); (D) S pielschmanm adult female. 15.1 mm (redrawn from Jones and Kumaran,
1964); (E) S. praematura larva, 3.6 mm (from Ozawa and Matsui. 1979); and (¥) S praematura adult female. 20.1 mm (redrawn from Jones
and Kumaran, 1964).
Development
Eggs
Although ovarian eggs are well-known for both species (Jones
and Kumaran. 1964; Sardou. 1974), the mode of spawning is
unknown. Watson and Leis (1974) reported planktonic Schind-
leria sp. eggs which they suggested were either pelagic or perhaps
dermersal eggs extruded in the net. The largest ovarian eggs lack
oil droplets and are irregular in shape, 0.35-0.40 mm in di-
ameter (S. praematura). or oval. 0.30 x 0.65 mm (S. pietsch-
manni'). Hydrated, planktonic eggs of Schmdlena sp. are oval,
554
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
0.50 X 1. 30 mm, contain no droplets, and have an unsculptured
chorion with a cap-Uke structure at one end. Incubation time
is not known.
Larvae
Morphology. — Larval size and degree of development at hatch-
ing are unknown. However, 5. pietschmanm at 1 .9 mm NL has
a rather large yolk sac (containing an apparently segmented yolk)
in addition to pigmented eyes and an open, presumably func-
tional, mouth. Notochord flexion occurs after 2.7 mm but before
3.5 mm NL in 5. pietschmanni, and before 4.3 mm in S. prae-
matura. Development to the essentially larval mature form is
gradual. The juvenile stage may be taken to begin with com-
pletion of the dorsal and anal fins and the acquisition of the
principal caudal rays (ca. 4-5 mm), and the adult stage to begin
when the male genital papilla or the ovaries of the female be-
come discemable (longer than ca. 9 mm SL). The distinctive
schindleriid terminal section at the rear of the vertebral column
does not become apparent until the late larval or early juvenile
period.
Aside from fin development, morphology changes little during
larval development. The swim bladder moves posteriorly from
myomeres 6-8 to myomeres 14-1 5 in 5. pietschmanni; a similar
migration presumably occurs in S. praematura (e.g., Sardou,
1974). Preanal length is greater in S. praematura than in S.
pietschmanm.
Pigmentation. — Sdn'mdXmids are lightly pigmented throughout
development (e.g.. Miller etal., 1979; Ozawa and Matsui, 1979).
During the larval and early juvenile period, S. pietschmanm
has one to four pairs of melanophores along the sides of the gut
(usually two or three pairs), one to four melanophores along the
ventral midline of the tail (usually two or three), and pigment
on the posterior dorsal surface of the swim bladder. The pos-
terior tail melanophore is typically more elongate than the others
(Fig. 298). All but the swim bladder pigment is lost during the
juvenile stage. Larval pigmentation of 5. praematura. as shown
by Ozawa and Matsui (1979), and juvenile pigment, shown by
Sardou (1974), are very similar to that of 5. pietschmanni. Like
S. pietschmanni, S. praematura retains only the posterior swim
bladder pigment in the adult stage (Fig. 298).
A/m5/(C5. — Meristics for Schindleria are: Vertebrae 15-25 -I-
12-21 = 33-44; D 15-22; A 10-14; P 15-17; and C 13prin. A
combination of caudal vertebrae and anal fin ray counts usually
will distinguish the two species.
The caudal fin rays are the first to develop, followed by the
dorsal and anal fin rays (forming simultaneously). Pectoral fin
rays are the last to ossify. Pelvic fins never form.
Relationships
Early life history characters, to the extent that they are pres-
ently known, do little to clarify the phylogenetic position of the
Schindlerioidei. For example, Gosline ( 1 963b, 1971) speculated
that Schindlerioidei might be derived from an ammodytoid
ancestor; however, while both suborders share some characters
(e.g., an elongate larval form with preanal length just over 50%
body length), they differ in other important ways (e.g., late de-
velopment of pectoral fin rays in schlindleriids and early de-
velopment in ammodytoids). Knowledge of spawning and early
development might aid in ascertaining schindleriid relationships
although at present this group seems destined to remain an
enigma.
(W.W.) Marine Ecological Consultants, 531 Encinitas
Boulevard, Suite 1 10, Encinitas, California 92024; (E.
G.S.) National Marine Fisheries Service, Southwest
Fisheries Center, PO Box 271, La Jolla, California
92038; (A. CM.) National Marine Fisheries Service,
Northwest and Alaska Fisheries Center, 2725
Montlake Boulevard East, Seattle, Washington 98 112.
Trachinoidea: Development and Relationships
W. Watson, A. C. Matarese and E. G. Stevens
THE blennioid infraorder Trachinoidea, as used here, con-
tains about 140 species in 1 1 families of morphologically
quite diverse, but generally small, primarily shallow-living tem-
perate and tropical marine demersal or burrowing fishes (Chias-
modontidae is bathypelagic; Cheimarrhichthyidae inhabits fresh
water). These families have not always been considered as closely
related (e.g., Gosline, 1968, 1971), but we follow Nelson (1976)
in considering them together here. Nelson ( 1 976) originally placed
16 families in the Trachinoidea, but subsequently synonymized
the Limnichthyidae with Creediidae (Nelson. 1978). Springer
(1978) removed Oxudercidae to the Gobiidae. Three other fam-
ilies are treated elsewhere in this volume; Bathymasteridae and
Dactyloscopidae with the Blennioidea (Matarese et al., this vol-
ume) and Opistognathidae with the Percoidei (G. D. Johnson).
In this brief review, we summarize the present state of knowl-
edge of the early life histories of trachinoid fishes and attempt
to determine whether such information contributes to our un-
derstanding of their phylogenetic relationships. Unfortunately,
early life histories, mostly incomplete, are known for only a
small number of species (Table 134). This paucity of early life
history data makes generalizations about development tenuous
at best, but for purposes of this paper the known taxa are con-
sidered representative.
Development
Eggs
Eggs are unknown for the Percophididae, Trichonotidae, and
Leptoscopidae. Only ovarian eggs have been described for the
WATSON ET AL.: TRACHINOIDEA
555
Table 134. Summary of Early Life History Information Available for Trachinoid Fishes.
Approxi-
Number
mate
Descriptions
llluslralions
of
niimh«*r nf
Family
genera
IJ Ul J 1 t-n^ t ^Jl
species
Dislnbution
Genera
Species
Genera
Species
Early life history source
Trichodontidae
2
2
North Pacific
1
1
1
1
Breder and Rosen, 1966; Marliave,
1981
Cham psodontidae
1
10
Indo-Pacific
I
1
1
I
Mito, 1962d, 1966
Chiasmodontidae
4
23
Worldwide (temperate
and tropical, marine)
1
1
1
1
Ahlslrom, pers. comm.; Lavenberg,
pers. comm.
Percophididae
6
17
Atlantic, Indo-Pacific
1
1
1
1
Crossland, 1982
Mugiloididae
3
36
Atlantic, Indian, Pacific
(subtropical and tropi-
cal)
1
5
1
4
Leis and Rennis. 1983; Mito, 1966;
Robertson, 1973, 1975a; Watson,
unpubl.
Trichonotidae
2
5
Indo-Pacific
1
2
1
1
Leisand Rennis, 1983
Cheimarrhichthyidae
1
1
New Zealand (freshwa-
ter)
0
0
0
0
McDowall, 1973c
Creediidae
7
14
Indo-Pacific
2
2
2
3
Leis, 1982; Leis and Rennis, 1983;
Regan, 1916; Watson and Leis,
1974
Trachinidae
1
4-5
Eastern Atlantic, Medi-
terranean
1
4
1
->
Breder and Rosen, 1966; Dekhnik.
1973; Ehrenbaum, 1905-1909; Mari-
naro, 1971; Padoa, 1956g; Russell,
1976; Schnakenbeck, 1928; Vod-
vanitskv and Kazanova, 1954
Uranoscopidae
8
25
Atlantic, Indian, Pacific
(shallow temperate
and tropical)
3
4
2
3
Dekhnik, 1973; Fritzsche, 1978; Mito,
1966; Robertson, 1974
Leptoscopidae
2
3
Australia, New Zealand
(marine)
0
0
0
0
Cheimarrhichthyidae (McDowall, 1973c). Six of the seven re-
maining families spawn small to moderate (0.70-2.45 mm di-
ameter), spherical, single pelagic eggs (Table 135). McDowall
(1973c) suggested a pelagic spawning mode for Cheimar-
rhichthyidae as well, unusual for the suggested riparian spawn-
ing habitat but consistent with the close relationship, or identity,
of Cheimarrhichthyidae with Mugiloididae. All pelagic eggs have
oil droplets (most have only one, 0.16-0.26 mm in diameter)
and all except some Uranoscopidae have smooth, unsculptured
chorions. Incubation periods range from 2 to 6 days and larvae
are not well developed at hatching (Trachinidae are somewhat
better developed, with pigmented eyes and pelvic buds).
Demersal egg masses (750-1,000 eggs) are produced only by
the Trichodontidae (Table 135). These eggs are large (3.52 mm
in diameter), slightly flattened, with an unsculptured chorion
and no oil droplet. Incubation is estimated at about one year
(Mariiave. 1981) and larvae are well developed at hatching.
Larvae
Larval stages are unknown for the Cheimarrhichthyidae and
Leptoscopidae. The described trachinoid larvae display only a
few unifying characteristics: (1) all are pelagic, hatching at ca.
2-15 mm (Table 1 36); (2) they pass through no specialized stages
(except the gargaropteron juvenile stage of the chiasmodontid
genus Kali): and (3) they metamorphose gradually to the de-
mersal juvenile stage at a small to moderate size (ca. 10-60
mm).
Mo/7)/)o/og)'. — Morphology is quite variable; however, larvae
are either relatively long and slender (Fig. 299: Trichodontidae,
Chiasmodontidae, Percophididae, Trichonotidae, Creediidae)
or rather robust (Fig. 300: Champsodontidae, Mugiloididae,
Trachinidae, Uranoscopidae). All the robust larvae and one of
the slender types (Trichodontidae) have somewhat rounded heads
with relatively short snouts. Preanal length in both types usually
is not more than 50% of standard length (60% or more in Cree-
diidae and Trichonotidae) and changes little during develop-
ment. Head and body spination are extremely variable. Pre-
opercular spination is known for six families: Trichodontidae,
Chiasmodontidae, Champsodontidae, Mugiloididae, Creedi-
idae, and Trachinidae. Champsodontid larvae develop a serrate
crest on the snout and head during the postflexion period, and
chiasmodontid larvae (except Kali: R. J. Lavenberg, pers. comm.)
develop small body spicules (Fig. 299) just before or during
notochord flexion.
Pigmenlalion. — Pigmenlalion of trachinoid larvae is quite vari-
able, from nearly absent to quite intense (Table 137). Larval
champsodontids, mugiloidids, trichonotids, and creediids re-
main lightly pigmented throughout development, while larval
trichodontids, chiasmodontids. trachinids, and uranoscopids
may become rather heavily pigmented. Pigmentation usually
increases with increasing larval size; trichonotids and creediids
change little in pigmentation with growth.
Head. — Eyes are pigmented at hatching for the demersally-
spawned Tnchodontidae, and for two of the six families with
pelagic eggs (Table 137). Pigmentation is present at hatching,
or subsequently develops, over the brain in five families. The
degree of pigmentation of other areas of the head is variable.
Gut. — Pigmentation typically is present dorsally over the gut
and swim bladder throughout larval development (absent only
in creediids and postflexion trichonotids). Other gut pigment is
variable.
556
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 135. Characteristics of Trachinoid Eggs.
Pelagic (P)
Egg
Oil droplets
or
Single
diameter
number: size
Attachments or
Incubation
Family
demersal (D)
or mass
(mm)
range (mm)
ornamentation
Pigmentation
penod
Source
Trichodontidae
D
Mass 750-
1,000
eggs
3.52
0
None
Amber
2 mo.-l
yr-
Breder and Rosen,
1966; Mariiave,
1981
Champsodontidae
P
Single
1.09-1.19
1: 0.17-0.22
None
Melano-
phores on
embryo
and oil
droplet
McDowell, 1973c;
Mito, 1966
Chiasmodontidae
P
Single
1.08-1.14
1: 0.26
None
Chorion rose
to amber
Ahlstrom. pers.
comm.
Percophididae
Unknown
Mugiloididae
P
Single
0.77-1.25
1: 0.16-0.25
None
Melano-
phores on
embryo
and oil
droplet
5-6 days
Mito, 1966; Rob-
ertson, 1973,
1975a
Trichonotidae
Unknown
Cheimarrhichthyidae
P (assumed)
Single
(ovarian)
McDowell. 1973c
Creediidae
P
Single
0.70-1.10
400-600 in
8-12 clus-
ters; co-
alesce to
3-8: 0.05-
0.10
None
Melano-
phores on
embryo
2 days
Lets, 1982; Watson
andLeis, 1974
Trachinidae
P
Single
0.94-1.37
1-30, co-
alesce:
0.19-O.25
None
4-5 days
Breder and Rosen,
1966; Dekhnik,
1973; Marinaro,
1971; Padoa,
1956g; Russell,
1976
Uranoscopidae
P
Single
1.52-2.45
3-27: 0.02-
0.15
Polygonal
network
on cho-
rion in
Uranosco-
pus
Melano-
phores on
yolk and
embryo
Dekhnik, 1973;
Fritzsche, 1978;
Mito, 1966; Rob-
ertson, 1975a
Leptoscopidae
Unknown
Trunk and tail. — Most trachinoid larvae display some degree
of pigmentation along the ventral margin of the tail (absent in
some mugiloididsand preflexion trachinids). Pigmentation (typ-
ically rather light) occurs along the dorsal margin of the trunk
and tail at some time during larval development in many tra-
chinoids. Internal pigment may develop above and below the
vertebral column (e.g., Trichodontidae).
Hypural margin. — Hypural pigment typically is light or absent
although its presence as a bar is diagnostic for the Trichodon-
tidae.
Fins. — Fins typically are unpigmented in trachinoid larvae, al-
though for some groups fin pigmentation can be diagnostic (e.g.,
the caudal and posterior dorsal and anal fin pigment of Trich-
Table 136. Size (mm SL) of Trachinoid Larvae at Selected Developmental Stages.
Prejuvenile or
Family
Hatching
Nolochord flexion
specialized stages
Juvenile
Trichodontidae
14.5
Before hatching
None
32-60
Champsodontidae
3.4-3.7
4.6-5.0
None
9.6-10.7
Chiasmodontidae
ca. 4
Before ca. 9
ca. 45
ca. 12-45
Percophididae
<16.0
Mugiloididae
2.2-3.0
3.7-4.8
None
10.0 to >12.6
Trichonotidae
5.2-6.3
None
>18.8
Cheimarrhichthyidae
s25
Creediidae
2.6-3.5
7.0-10.2
None
>11.0, S29.2
Trachinidae
3.2
5.0-10.0
None
13-15
Uranoscopidae
>2.5-4.38
None
£23
Leptoscopidae
No information
WATSON ET AL.: TRACHINOIDEA
557
B
^^S^
D
Fig. 299. (A) Trichodontidae: Trichodon trichodon. 13.0 mm, from Marliave (1981); (B) Chiasmodontidae: Pseitdoscopelus sp.. 14.0 mm,
CalCOFI station 5710-5-130.80 (approximately 24°49'N, 1 16°49'W); (C) Percophididae: Hemerocoeles sp., 16.0 mm, redrawn from Crossland
(1982); (D) Tnchonotidae: Tnchonotus sp., 5.9 mm, from Leis and Rennis (1983); and (E) Creediidae: Limmchlhys donaldsom. 1 1.0 mm, from
Leis(1982).
onotidae or the early developing heavily pigmented pelvic fins
of trachinids),
Meristic characters— WtrXthraX and fin ray counts are summa-
rized in Table 138. The sequence of fin ray formation, incom-
pletely described for most families, appears to be quite variable
except that the caudal fin is first to begin ossification of rays in
all but the trachinids (the caudal is second in this family, fol-
lowing the pelvic fins). Dorsal and anal fin rays are second to
form in four families (Mugiloididae, Tnchonotidae. Creediidae,
Uranoscopidae), while pectoral fin rays are second in two (Trich-
odontidae and Chiasmodontidae) and pelvic fin rays in one
(Champsodontidae).
Special structures. Sipec\a.\ structures are generally lacking in
trachinoid larvae. Only the elongate opercular appendage of
558
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 137. Summary of Pigmentatton (Melanin Only) of Larval Trachinoid Fishes. Key: +, present; 0, absent; J, increasing with devel-
opment; 1, decreasing with development; 0-+, initially unpigmented, becommg pigmented with development; An. anterior; Po. posterior.
Eye at
hatching
Head
Gut
Trunk
and tail
Ventra
margin
Family
Brain
Jaws
Snout
Opercle
Isthmus
Nape
An-
terior
Dor-
sal
Ven-
tral
Ut-
eral
Pre-
flexion
Flexion
Trichodontidae
+
+1
+
+
0
0
0- +
0
0
0
-1-
+
Champsodontidae
+
+
0
+ 1
0
0
0- +
0
-1-
0- +
+
-1-
-(-
Chiasmodontidae
0
+
+
0
0
0- +
0
0
-1-
0
0
4-
+
Percophididae
Mugiloididae
0
0
0, or
+1, or
+T
0
0
0
0, or
+ 1
+
0
0
0
0
0
0
0, or
+
-1-
+ T
0
0, or
+
0
0- +
0, or
+
+
Trichonotidae
Cheimarrhichthyidae'
Creediidae
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
+1
0
0
+
0
0
+ Po
+ Po
+ Po
+ Po
Trachinidae
+
0. or
+ 1
0, or
+T
0, or
+ T
0-+I
0
0, or
+
0
+
0
0- + I
0, or
+
+
Uranoscopidae
Leptoscopidae'
0
+ 1
0- +
+1
0- +
0- +
0- +
+
+T
0-+T
or
+ 1
0-+1
or
+ T
+ T. or
+ 1
+ T. or
+ 1
' Larvae unknown
Table 138. Selected Meristics of Trachinoid Fishes.
Family
Dorsal fin
Pectoral
fin
Pelvic
fin
Pnmary
caudal
fin rays
Trichodontidae X-XVI + O-I, 13-19
Champsodontidae V + I, 1 8-22
Chiasmodontidae VI-XIII + 18-28 0-1,17-28 10-15 1,5
,27-31
21-23
I, 5
12-15
12-15 + 34-40
47-50
, 17-20
9-13
1,5
10 + 19-22 =
29-32
17
33-44
Marliave, 1981; NWAFC,
unpubl.
de Beaufort and Chap-
man, 1951; Matsubara
et al., 1964; Milo, 1962d
Johnson and Cohen.
1974; Lavenberg, pers.
comm.; Norman,
1929
Percophididae
0-IX + 14-31
O-I,
15-42
20-28
I. 5
15
8-9 -1- 19-21 =
27-30
Ginsburg, 1955; Iwamoto
andStaiger, 1976; Mil-
ler and Jorgenson, 1973
Mugiloididae
IV-VII. 19-28
1
16-26
15-22
I, 5
14-15
10-16 + 18-22 =
28-38
Cantwell, 1964
Trichonotidae
III-VII, 40-46
36-40
12-14
I, 5
13
15 + 40 = 55
Leisand Rennis, 1983;
Masudaet al., 1975
Cheimarrhichthyidae
IV-VI, 18-21
I-II
14-16
14-18
I, 5
12-15
12 + 20-21 =
32-33
McDowall, 1973c
Creediidae
18-40
25-40
11-17
None,
or I,
3-5
10
37-59
Leisand Rennis, 1983;
Smith, 1961
Trachinidae
V-VII + 21-32
25-36
15
1,5
14
11-12 + 23-31 =
34-43
Padoa, 1956g; Russell,
1976
Uranoscopidae
0-V + 12-19
O-I
12-19
13-24
1,5
11-14
9-12 + 14-17 =
25-29
Berry and Anderson,
1 96 1; Fritzsche, 1978;
Marshall, 1965; Miller
and Jorgenson, 1973;
Mito, 1966; Scott et al.,
1974; Smith, 1961;
Wade, 1946
LeptoscopicJae
34-35
37
I, 5
lO-H
Gosline, 1968; Scott et
al., 1974
WATSON ET AL.: TRACHINOIDEA
Table 137. Extended.
559
Trunk and rail
Venlral
margin
Dorsal margin
Latera!
Fins
Post-
Pre-
Post-
Prc-
Post-
Hypural
Pec-
flexion
flexion
Flexion
flexion
flexion
Rexion
flexion
margin
Dorsal
Anal
toral
Pelvic
Caudal
Source
+
0- +
An
+ 1
0
+ T
+
0- +
0
0--h
0
0-4-
Marliave, 1981
+
+
+
+
0
0, or
-1-
0, or
+
0
+ 1
+1
0
0
+ 1
Mito, 1962d, 1966
+
0, or
+
0, or
+
+
0
0
0
0
0
0
0
0
0
Ahlstrom, pers. comm.;
Lavenberg, pers.
comm.
0
0
0
0
0
0
0
0
McDowall, 1973c
+ , or
+ 1 Po
0, or
0
0, or
0, or
0. or
0
0
0
0, or
0, or
0
Leisand Rennis, 1983;
+ 1
+ lPo
+ Po
+ Po
+ An
+
-(-
Mito, 1966; Robert-
son, 1973; Watson,
unpubl.
+ Po
+ Po
+ Po
-1- Po
-1- Po
+ Po
+ Po
0
+ Po
+ Po
0
0
+
Leisand Rennis, 1983
+ Po
+ Po
+ Po
0, or
+ Po
0
0
-1-
0--h
+ lPo
+ iPo
0
0
+
Leis, 1982; Leisand
Rennis, 1983; Regan,
1916
+
0, or
+ An
0. or
+ An
0, or
-1- An
0
0, or
+ An
-1- An
+
0-+T
0
0
-1-
0
Ehrenbaum, 1905-1909;
Padoa, 1956g; Rus-
sell, 1976
0, or
+
0
0
4-
+ An
-1-
-1-
0, or
0, or
0
0
0
Dekhnik, 1973; Mito,
+
+1
+1
1966; Pearson, 1941
Champsodon. and the body spicules of the chiasmodontids (ex-
cept Kali) are distinctive. Trachinus vipera has precocious, en-
larged, and heavily pigmented pelvic fins.
Relationships
The trachinoid families summarized here are presumed to be
deinved from the Percoidei, or in some cases to belong them-
selves to the Percoidei (e.g., Trichodontidae, Champsodontidae,
Chiasmodontidae: Gosline, 1971). Therefore, in the following
discussion we consider early life history characters shared with
the Percoidei as primitive. Characters shared with other Blen-
nioidei are, somewhat arbitrarily, considered to be derived. Our
purpose in classifying characters into these categories is not to
develop a new phytogeny of the Trachinoidea based on early
life history, since far too little in known to allow such an un-
dertaking, but rather to determine whether such characters sup-
port our treatment of the Trachinoidea as a monophyletic group.
Six of the 1 1 trachinoid families retain the pelagic spawning
Table 1 39. Summary of Early Life History Characteristics of the Trachinoidea. The percoid condition is assumed to be primitive, while
the blennioid condition is assumed to be derived. The percoid condition includes spawning of pelagic eggs which soon hatch to poorly-differentiated
larvae, a moderately deep body, myomeres mid-to-upper twenties, development of dorsal and anal fin rays before pectoral and/or pelvic fin rays,
and five pelvic fin rays. The blennioid condition includes spawning of non-pelagic eggs with an extended incubation period and hatching of well
developed larvae having pigmented eyes, an elongate shape, myomeres thirty or more, development of pectoral and^'or pelvic fin rays before
dorsal and anal fin rays, and fewer than five pelvic fin rays. It should be understood that spawning mode, incubation period and development at
hatching tend to be correlated, as are larval shape and number of myomeres.
Eggs
Larvae
Urval
shape
Number of
myomeres
Pectoral
pelvic fin
ossification
Reduced
number of
pelvic rays
Pre- •
opercular
spination
Larval
Family
Spawning
mode
Incubation
penod
Development
at hatching
pigmen-
tation
Trichodontidae
Champsodontidae
Chiasmodontidae
Percophididae
Mugiloididae
Tnchonotidae
Cheimarrhichthyidae
Creediidae
Trachinidae
Uranoscopidae
Leptoscopidae
Derived
Primitive
Primitive
Primitive
Primitive?
Primitive
Primitive
Primitive
Derived
Derived
Primitive
Derived
Derived
Derived
Primitive
Primitive
Primitive
Derived
Primitive
Derived
Primitive
Derived
Derived
Primitive
Derived
Derived
Primitive
Primitive
Derived
Primitive
Derived
Primitive
Primitive
Derived
Derived
Derived
Primitive
Derived
Derived
Derived
Primitive
Primitive
Primitive
Primitive
Derived
Primitive
Primitive
Primitive
Primitive
Primitive
Primitive
Primitive
Derived
Primitive
Primitive
Present
Present
Present
Absent
Present
Absent
Present
Present
Absent
Heavy
Light
Heavy
Light
Light
Light
Light
Heavy
Heavy
560
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
B
Fig. 300. (A) Champsodontidae: Champsodon snyderi. 9.6 mm, from Mito (1962a); (B) Mugiloididae: Parapercis schauinslandi. 5.3 mm, Kahe
Point, Oahu, Hawaii (approximately 21°16'N, ISS'S'W); (C) Trachinidae: Trachinus vipera. 7.5 mm, redrawn from Schnakenbeck (1928); and
(D) Uranoscopidae: Astroscopus giMatus. 4.9 mm, from Pearson (1941).
WATSON ET AL.: TRACHINOIDEA
561
mode (Tables 135 and 139) typical of the mainne percoids, one
shares with the other Blennioidei the condition of spawning non-
pelagic egg masses. Among the pelagic spawners, four retain the
percoid-like condition of early hatching of poorly-differentiated
larvae; two share with the demersal spawners the condition of
a relatively long incubation and hatching of well developed lar-
vae with pigmented eyes.
The larvae of four families are moderately deep-bodied, a
character shared with the majority of percoids. Each of these
families (except Trachinidac) contains at least some species with
myomeres numbering in the mid-to-upper twenties: typical per-
coid counts. Five trachinoid families resemble blennioids in
having elongate larvae, usually with more than 30 myomeres.
All trachinoid larvae (except some Trachinidae) follow the
typical perciform pattern of beginning caudal fin ossification
first; lai^ae of five families follow the percoid pattern of begin-
ning ossification of dorsal and anal fin rays before pectoral and
pelvic fin rays. Four families share with the other blennioids
the early acquisition of pectoral and/or pelvic fin rays. All tra-
chinoid families share with the other blennioids the jugular
placement of pelvic fins, but only one family (not all species)
also shares the blennioid condition of fewer than five pelvic fin
rays.
Larval pigmentation and preopercular spination of the Tra-
chinoidea (Table 139) are difficult to assess, since both range
from absent to highly developed in both the Percoidei and Blen-
nioidei. The distribution of these characters is listed in Table
1 39 to aid in determining relationships among the Trachinoidca.
Based solely on early life history characters (Table 1 39), the
Uranoscopidae and Mugiloididae (including Cheimarrhichthyi-
dae?) appear to be the most percoid-like members of the Tra-
chinoidca, while Trichodontidae are most like the other Blen-
nioidei. Two points become clear in considering the contribution
of early life history to the understanding of trachinoid phylog-
eny: (1) the Trachinoidea is a very diverse, probably polyphy-
letic, group; and (2) much more early life history data are needed
before any substantial contribution can be made to the under-
standing of this group.
(W.W.) Marine Ecological Consultant.s. 531 Encinitas
BouLEVARtJ, Suite 110, Encinitas, California 92024;
(A.C.M.) National Marine Fisheries Service, North-
west AND Alaska Fisheries Center, 2725 Montlake
Boulevard Ea.st, Seattle, Washington 98112; (E.G.S.)
National Marine Fisheries Service, Southwest Fisheries
Center, PO Box 271, La Jolla, California 92038.
Notothenioidea: Development and Relationships
E. G. Stevens, W. Watson and A. C. Matarese
NOTOTHENIOIDEA comprises 5 families with 35 genera
and about 100 species (Table 140). These familes are
endemic to the Antarctic and Subantarctic regions (DeWitt,
1971; Norman, 1938a; Wyanski and Targett. 1981). Adults,
ranging from 100 to 900 mm SL, occupy several habitats from
the surface to several hundred meters depth and are often as-
sociated with continental and island slopes and shelves. Some
species are adapted for living close to the undersurface of ice.
Discussions of the systematic position of notothenioids are
found in Gosline (1968) and Norman (1938a), who consider
them Perciformes or perciform relatives on the basis of the
adult cranial osteology; the jugular position of the pelvic fins,
which have one spine and five rays; and the caudal fin ray
number, usually 14. Both note the reduced number of pectoral
radials found in Notothenioidea. Gosline (1968) unites the no-
tothenioids with trachinoids and blennioids using characters
such as the one to one ratio of vertebrae to dorsal and anal fin
rays, more than 2 5 vertebrae, and fewer than 1 5 branched caudal
rays. Gosline ( 1 968), Norman ( 1 938a), and other recent workers
(i.e., Andersen and Hureau, 1979) separate Nototheniidae and
Harpagiferidae making a total of five families (this classification
is used here), whereas Nelson (1976) follows Berg (1940) and
Table 141. Notothenioidea: Egg Diameter (mm) and Larval Size
at Selected Developmental Stages (mm SL).
Egg
Nolochord
Family
diamcler
Hatchmg
flexion
Juvenile
Bovichthyidae
Unknown
Unknown
Unknown
ca. 25
Nototheniidae
1.2-4.0
6-14
9-20
25-60
Harpagifendae
2.4-3.0
7-13
ca. 9-13
35-38
Bathydraconidae
1.5-3.0
Unknown
18-24
24-34 +
Channichthyidae
2.8-4.5
ca. 14
18-42
50-60 +
Table 140. Notothenioidea: General Summary and Early Life History Information.
Number of
genera
Approximate
number of
species
Distnbulion
Ear
y life history
Descnplions
Illustrations
Famil\
Genera
Species
Species
Bovichthyidae
Nototheniidae
Harpagiferidae
Bathydraconidae
Channichthyidae
3
8
5
10
U
12
50
15
14
17
Antarctic, Subantarctic
Antarctic, Subantarctic
Antarctic, Subantarctic
Antarctic
Antarctic
0
5
2
7
8
0
16
4
8
8
0
12
3
6
7
562
ONTOGE>fY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM
Fig. 301. Notothenioid larvae (from top to bottom): Nototheniidae: Patagonololhcn larsent. 35 mm. from North and While (1982); Harpa-
giferidae: Harpagifer bispinis, 18.2 mm, from Everson (1968); Harpagiferidae: Artedidraco inints. 24.0 mm. from Efremenko (1983); Bathydra-
conidae: Psilodraco breviceps. 16.9 mm, from Efremenko (1983); and Channichthyidae: Pagetopsis macroplerus, 19 mm, redrawn by H. Orr from
Regan (1916).
STEVENS ET AL.: NOTOTHENIOIDEA
563
Table 142. Notothenioidea: Selected Meristics. Sources listed here do not include the following cited in the text: Andriashev (1959); DeWitt
(1970); Norman (1937, 1938); Nybelin (1947, 1951); Regan (1913d. 1916); Yefremenko (1979a, b). Omissions indicate no data were found in
the literature. While this paper was in press a revision of the Nototheniidae was published (Anderson, 1984).
Fin rays
Vertebrae
Number _
of
Family
Pnncipal Pre-
Genus
species
Di
D;
A
Pec.
Pel.
caudai caudal
Caudal
Tolal
Sources
Bovichthyidae
Bovichlhys
10
VII-VIII
18-21
13-19
14-16
1.4-5
41
(Isp.)
Hureau and
Tomo, 1977
Colloperca
1
VI-VIII
21-24
20-24
17
Pseudophrites
1-2
VII-VIII
19-20
23-25
18
1.5
Scott, 1962
Nototheniidae
Dissostichus
2
VIII-X
25-29
25-30
25-29
1,5
18-20
54-55
de Ciechomski
and Weiss,
1976
Eleginops
1
VIII-IX
24-26
22-24
22-24
1,5
12
30-31
42-43
de Ciechomski
and Weiss,
1976
Pleuragramma
1
Vl-VIII
33-38
36-39
20-21
1,5
18-20
53-55
Aelhotaxis
1
VII
34
30
27-28
1.5
16
36
52
DeWitt, 1962b
Cryolhenia
1
IV-VI
34-36
33-35
24-26
1,5
14-17 13-18
33-37
50-53
Daniels, 1981
Pagothenia
2
IV-VI
30-37
26-33
23-24
1,5
14
DeWitt, 1964b
Palagonololhen
16
IV-VIII
27-38
27-37
24-28
1.5
54-55
(Isp.)
Notothenia
ca. 27
III-VIII
25-41
21-36
16-30
1,5
14-16 14-22
27-37
44-58
Harpagiferidae
Artedidraco
5
II-IV
24-30
16-21
14-17
1.5
Dollwdraco
1
II-III
22-25
14-15
16-18
1.5
13
21-22
34-35
DeWitt and
Tyler, 1960
Histiodraco
1
11-111
26
17
18-19
1,5
Pogonophryne
7
11-111
25
16-18
19 (Isp.)
1,5
15
23
38 (Isp.
)
Harpagtfer
1
Ill-V
21-26
16-21
15-18
1,5
10-11
24-25
35-36
Bathydraconidae
Balhydraco
4
0
34-40
29-31
-)T
I, 5
52-53
Gerlachea
1
0
45-47
34-35
26-28
1,5
25
38
63
Racoviizia
1
0
30-37
27-31
22-25
1.5
19-20
32-34
52-54
DeWitt. 1964a
Phonodraco
I
0
34-38
29-33
21-22
16
33-34
49-50
Cygnodraco
1
0
61-66
35-38
22-23
Parachaenichthys
2
0
42-46
30-33
21-23
Psilodraco
1
0
27-30
27-29
25-27
16
48
Gymnodraco
1
0
28-30
22-26
21-22
I, 5
13 20-21
28
48-50
DeWitt and
Tyler. 1960
Akarota.xis
1
0
31-33
27-28
22-23
1.5
12-13 17
33
50
DeWitt and
Hureau, 1979
Vomeridens
1
0
34
32
31-33
1,5
12 18
36
54
DeWitt and
Hureau, 1979;
DeWitt, 1964a
Channichthyidae
Champsocephalus
2
VII-XI
32^1
31-tO
22-29
1,5
16-18
60-64
Kock, 1981
Pagetopsis
2
IX-XV
26-31
24-27
22-23
1,5
18-20
56-58
Pseudochaenichthys
VIII-XI
28-32
27-31
22-25
1.5
16-18
51-55
Kock, 1981
Neopagetopsis
XIV
34
29
26
Dacodraco
III
32
29
24
1,5
Chaemchthys
VI-IX
30-35
27-34
18-22
1,5
54-56
Chaenocephalus
V-IX
37-42
36-40
23-26
1,5
18-20
60-64
Kock. 1981
Cryodraco
III-V
40-45
39-46
23-25
1,5
20-22
63-65
Chionodraco
V-VII
37-51
33-38
21-24
1,5
16-18
60-64
Chaenodraco
VI-VIII
38-42
32-35
22-25
1,4
18-20
60-62
Chionobathyscus
564
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Greenwood et al. ( 1 966) in placing Harpagiferinae as a subfam-
ily of Nototheniidae (Table 140).
Development
The first work on the early life history of these fishes was
undertaken with material collected on the polar expeditions of
the early 20th century. Regan (1916) illustrated larvae of seven
notothenioid species. Additional early life history data were
sparse until the Antarctic expeditions of the second half of the
century. In the last 30 years efforts have been directed toward
understanding the biology, ecology, population dynamics, and
physiological adaptations of these fish. In these investigations
some early life history data have been acquired. Larvae of 36
species have been described (Table 140). The most comprehen-
sive summaries are the key in North and White (1982), the
papers of Yefremenko (1979b, c) and the atlas of Efremenko
(1983a).' No early life history data are available for the family
Bovichthyidae except for a brief description of the behavior and
an illustration of a 25 mm prejuvenile of Bovichthys variegatus
(Robertson and Mito, 1979).
Eggs
Eggs of four notothenioid families have been described in-
cluding some known only from studies of ovaries (Table 141).
Eggs are moderate to large (1.2-4.5 mm diameter) with large
yolks, no oil globules, and small perivitelline spaces (Marshall,
1953; Andriashev, 1965; Dearborn, 1965). In one species, No-
tothenia ( Trematomus) hernacchii. eggs are bright yellow to deep
brown. Most species are demersal spawners; nesting behavior
has been observed in N. hernacchii (Moreno, 1980) and Har-
pagifer hispinis (Daniels, 1978). However, Notothenia micro-
lepidola eggs have been reported from plankton collections
(Robertson, 1975a). The demersal eggs are sticky, clinging to
substrate or algae. One species, N. neglecta. reared in the lab-
oratory from artificially fertilized eggs, has an incubation time
of 103-150 days and hatches with a well-developed, heavily
pigmented body, black eyes, and a large yolk sac (White et al.,
1982). Daniels (1978) reports an incubation time of 14 to 18
weeks for H. hispinis.
Larvae
Morphology— The described larvae of 36 species display some
morphological similarity (snout-anus length), and some diver-
sity (snout length and body shape). Preflexion larvae, ca. 6-18
mm SL, are elongate with large pectoral fins and moderate to
wide finfolds. Channichthyid larvae have well developed pelvic
fins at this stage and more elongate snouts than larvae of other
notothenioids. Some species have large yolk sacs which persist
after notochord flexion has begun. Preanal lengths range from
slightly under to slightly over 50% of body length.
Dunng flexion and postflexion stages most larvae maintain
their elongate shape (Fig. 301). However the larvae of the har-
pagiferid genera Artedidraco and Pogonophryne become very
robust (North and White, 1982; Efremenko, 1983a). Notochord
flexion occurs between 9 and 42 mm with larval Harpagiferidae
and Nototheniidae flexing at the shortest lengths, Channichthyi-
dae at the longest, and Bathydraconidae at intermediate lengths
(Table 141). Size at transformation to the juvenile stage also
spans a wide range with Harpagifer hispinis settling at 18.3 mm
' Efremenko and Yefremenko are alternative transliterations of the
name of the same author.
(Everson, 1 968) and pelagic larvae of other species reaching 24-
60 mm (de Ciechomski and Weiss, 1976; North and White,
1982; Yefremenko, 1979b, c).
Larvae of all species develop pelvic fins. Channichthyid larvae
retain their elongate snouts and develop teeth and preopercular
and rostral spines not reported for other notothenioids (Fig.
301).
Pigmentation. — Pi%menx patterns of all known larvae are highly
specific and are useful identification criteria. The amount and
location of pigment varies within families and the amount usu-
ally increases with development. A few species have general
body pigment, but most exhibit patterns in one or more of the
following areas: dorsal body margin, ventral body margin, body
midline, peritoneum, gut, and along the myosepta. The occipital
and parietal areas typically are pigmented; many species have
snout, opercular, and jaw pigment as well. The paired fins are
usually pigmented. Pigment is found at the base of the caudal
fin in most species, associated with the posterior margin of the
hypural elements or the bases of the caudal rays.
Meristics. — Meristics are from counts given for adults by Regan
(1913d, 1916), Norman (1937, 1938a), Nybelin (1947, 1951),
Andriashev (1959), and DeWitt (1970) (Table 142). Vertebral
counts are especially useful diagnostic features within and be-
tween families. The dorsal, anal, pectoral, and vertebral counts
have been the most significant characters linking larvae to adults
(Yefremenko, 1979b, c). The sequence of fin formation is the
same in Nototheniidae, Harpagiferidae, and Bathydraconidae
with pectoral and caudal fins forming first, followed by pelvics,
with dorsal and anal last to ossify. In Channichthyidae the pel-
vics are precocious and are present in yolk-sac larvae.
Relationships
Knowledge of the early life history of Notothenioidea has not
contributed to understanding relationships between Blennioidei
and other perciform suborders, but does ofler some clues to
relationships within the suborder. The lengthy ovarian egg de-
velopment (Dearborn, 1965; Everson, 1970) is probably related
to the cold environment. In other aspects of spawning, i.e.,
nesting behavior, long incubation time, and laying of demersal
adhesive eggs, this infraorder resembles other cold-water blen-
nioids. The well developed state of newly hatched larvae and
the sequence of fin development as well as the general lack of
specialized larval structures are also blennioid features. Further
study of developmental characters, such as the sequence of os-
sification, might contribute to better understanding of the re-
lationships among the Blennioidei. Superficial morphological
and meristic resemblances exist among notothenioid larvae and
those of other blennioid species, for example, the notothenioid
Patagonotothen larscni (Fig. 301), the trachinoid Trichodon
trichodon (see Trachinoidea, this volume) and the blennioid
Heterostichus rostratus (see Blennioidea, this volume). As re-
lationships among the blennioids become better known, their
relationship with other perciforms might become clearer.
(E.G.S.) National Marine Fisheries Service, Southwest
Fisheries Center, PO Box 271, La Jolla, California
92038; (W.W.) Marine Ecological Consultants, 531
Encinitas Boulevard, Encinitas, California 92024;
(A. CM.) National Marine Fisheries Service, North-
west AND Alaska Fisheries Center, 2725 Montlake
Boulevard East, Seattle, Washington 981 12.
Blennioidea: Development and Relationships
A. C. Matarese, W. Watson and E. G. Stevens
THE Blennioidea is composed of 16 families with about 182
genera and 759 species (Table 143). The families discussed
here are those included in the infraorder Blennioidea by Nelson
(1976), as amended by the current literature. For convenience
we divide the infraorder into a tropical and a northern group.
The tropical group is similar to Gosline's ( 1 968) superfamily
Blennioidae except for the following: 1) Ophiclinidae and Per-
onedysidae are synonymized with the Clinidae (George and
Springer, 1980); 2) Dactyloscopidae is included (George and
Springer, 1980); and 3) Congrogadidae is excluded (Winterbot-
tom, 1982'). The northern group is similar to Gosline's (1968)
superfamily Zoarceoidae except that we include the Bathymas-
teridae (Anderson, 1984). The Zoarcidae is treated separately
(Anderson, this volume).
The majority of species (80%) belong to four tropical families:
Tripterygiidae, Clinidae, Labrisomidae, and Blenniidae. Of the
northern forms, only the family Stichaeidae represents a sig-
nificant percentage (8%) of the species. Tropical Blennioidea
inhabit primarily the Indo-West Pacific south to Australia, while
northern fishes inhabit the North Atlantic and North Pacific
(Table 143). Occasionally, representatives of mainly tropical
families occur in boreal waters (e.g., Clinidae and Blenniidae),
and members of northern families may occur in the subtropics.
Some dactyloscopids inhabit fresh water. Four families are
monotypic and three of these are endemic to the northeast Pa-
cific.
As a group most of the 16 families in Blennioidea are not
well understood, probably due to their lack of commercial im-
portance, small size and cryptic habits. In general, the tropical
and more speciose families (e.g., Blenniidae) are better known
than the northern families. Monotypic families are quite poorly
known. Although sparse and incomplete, some early life history
information is available for II of the 16 families (Table 143).
In most cases, however, the data on few species may not be
representative of the family. Among the families in the infraor-
der, the Blenniidae has the greatest number of species (22) de-
scribed; but with about 319 species in the family, this amounts
to fewer than 10%. Morphology, pigment, and meristics of lar-
vae in the infraorder are diverse (Figs. 302, 303).
Development
Eggs
Fishes in this infraorder spawn demersal eggs (Table 144),
except for some clinids. Clinids of the tribe Ophiclinini are
ovoviviparous (George and Springer, 1980), while those of the
tribe Clinini are viviparous (Penrith, 1969; Hoese, 1976).
Most blennioid eggs are spherical to somewhat flattened, pos-
sess one to several oil droplets, are attached to one another (and
often to a substrate) by filaments or other adhesions, and have
' Winterbottom, R. 1982. The perciform fish family Congrogadidae—
biogeography and evidence for monophyly. Amer. Soc. Ich. Herp., oral
paper, 62nd annual meeting.
a smooth unsculptured chorion. Sizes range from among the
smallest offish eggs (Blenniidae, 0.50 mm) to among the largest
(Anarhichadidae, 8.0 mm). Incubation periods range from 6 to
70 days. Eggs are unknown for four families: Xenocephalidae,
Ptilichthyidae, Zaproridae, and Scytalinidae.
Parental care is common among most families; e.g., in sti-
chaeids. males or females may guard egg masses (Shiogaki and
Dotsu, 1972a; Shiogaki, 1981, 1982). In an extreme example
of parental care, male dactyloscopids incubate eggs in ball-like
clusters carried beneath the pectoral fins (Dawson, 1982).
Larvae
Morphological characters— ^Xenmoidea larvae hatch at sizes
ranging from as small as 2.0 mm (Blenniidae) to as large as 1 7.0
mm (Anarhichadidae) (Table 145). Larvae of the northernmost
families hatch at more than twice the size of larvae of the more
tropical families (i.e., averaging ca. 1 1.5 mm versus ca. 4.5 mm).
Size at which notochord flexion is complete is also variable, but
tropical larvae are usually fully flexed by ca. 10.0 mm whereas
northern larvae do not complete flexion until ca. 20.0 mm. At
least three families have larvae with an extended pelagic exis-
tence: Blenniidae, Cryptacanthodidae, and Zaproridae. Mem-
bers of the blenniid tribe, Salariini, have the only well-docu-
mented, prejuvenile pelagic stage (Miller et al.. 1979; Leis and
Rennis, 1983). This has been termed the "ophiblennius"" stage
and usually occurs between 4.6 and 26.0 mm (Fig. 302). At least
two families, Cryptacanthodidae (Shiogaki, 1982) and Zapror-
idae (Haryu and Nishiyama, 1981), have heavily pigmented
larvae and prejuveniles that are extensively collected in surface
nets suggesting an extended pelagic existence (Fig. 303C, G).
Most blennioid larvae do not undergo a marked metamorphosis.
Transformation is usually complete in tropical forms by 26.0
mm, but may begin as early as 10.0 mm in some families (Trip-
terygiidae and Blenniidae). Larvae in the more northern families
transform at a slightly larger size, ca. 30.0-40.0 mm, although
Ptilichthys transforms at ca. 1 14.0 mm.
Among the tropical families, larval Tripterygiidae, Clinidae,
and Labrisomidae share many similar morphological features.
They are moderately elongate, have a preanal length about 50%
BL (slightly less in labrisomids), possess a large swimbladder,
and usually lack preopercular spines (Figs. 302A, B, C, D).
Heads are small, sometimes rounded, with a short snout. Mouths
extend just beyond the anterior eye margin. In tripterygiid and
clinid larvae, the gut is initially straight but coils during flexion.
The blenniids include many larval forms with diverse mor-
phological features. According to Leis and Rennis (1983), how-
ever, larvae are more similar within tribes than between tribes.
Most species are moderately elongate (Nemophini includes both
slender and robust forms), becoming either more slender (Nem-
ophini) or more robust (Salariini) with development. Heads are
short, rounded, and broad becoming more elongate with de-
velopment (except Salariini larvae in which the snout elongates
early in the preflexion stage). The gut is short to moderate (usu-
ally < 50% BL), and eventually coiled if not so initially. L-arval
preopercular spination may be elaborate: spines can be numer-
565
566
ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM
Table 143. General Summary and Early Life History (ELH) Information in Blennioidea.
Early life history
Tax on
Approx.
Number of number of
genera species
Primary Dislnbution
Number Number
of of
genera species
Number
of
species
illus-
trated
Pnmao' early life history sources
Blennioidea
Dactyloscopidae 6
Xenocephalidae 1
Notograptidae 2
Tripterygiidae 18-19
Clinidae 26
Chaenopsidae 1 0
Labrisomidae 1 4
Blenniidae S3
20 Atlantic, Pacific
(tropical)
1 New Ireland, New Guinea
3 Australia
75-95 Atlantic, Pacific.
Indian (tropical)
85 Atlantic, Pacific,
Australia
55
100
289-319 Indo-Pacific
Atlantic. Pacific
(tropical)
New World (tropical)
0
0
2
3
7
22
Bathymasteri
dae
3
7
North Pacific
2
3
Stichaeidae
37
54
North Atlantic, Pacific
14
18
Cryptacanthodidae
3
4
North Atlantic, Pacific
3
3
Pholidae
4
13
North Atlantic, Pacific
3
5
Anarhichadidae
2
6
North Atlantic, Pacific
2
3
Ptilichthyidae
I
1
Northeast Pacific
1
1
Zaproridae
1
1
Northeast Pacific
1
1
Scytalinidae
1
1
Northeast Pacific
0
0
Dawson 1982
7 Graham, 1939; Leis and Rennis, 1983;
Miller etal., 1979; Ruck, 1973a,
1980; Shiogaki and Dotsu, 1973;
Watson, unpubl.; Wirtz, 1978
14 Bamhart. 1932; Padoa, 1956h; Shio-
gaki and Dotsu, 1972b; Sparta,
1948; Stevens, unpubl.; Watson, un-
publ.
Bohlke, 1957; Stephens et al., 1966
3 Breder, 1939; Breder, 1941; Springer.
1958; Watson, unpubl.
27 Cipria, 1934. 1936; Dotsu. 1982; Dot-
su and Monuchi, 1980; Dotsu and
Oota, 1973; Dutt and Rao, 1960;
Eggert, 1932; Fishelson, 1963, 1976;
Fives, 1970a; Ford, 1922; Fritzsche,
1978; Hildebrand and Cable, 1938;
Lebour, 1927; Leis and Rennis.
1983; Lippson and Moran, 1974;
Miller etal,. 1979; Mito, 1954;
Munro, 1955; Peters. 1981; Qasim,
1956; Rao. 1970; Russell, 1976; Ste-
vens and Moser, 1982; Thomson and
Bennett. 1953; Watson, 1974. un-
publ.; Wickler. 1957
1 Breder and Rosen, 1966; Fitch and
Lavenberg, 1975; Matarese, unpubl.
15 Breder and Rosen. 1966; Faber, 1976;
Hart, 1973; Marliave, 1975; Ma-
tarese, unpubl.; Peppar, 1965; Rass,
1949; Russell. 1976; Shiogaki. 1981;
Shiogaki. 1983; Shiogaki and Dotsu,
1972a; Tokuya and Amaoka, 1980;
Wourms and Evans, 1974
2 Hart, 1973; Matarese, unpubl., Shio-
gaki, 1982
3 Breder and Rosen, 1 966; Marliave,
1975; Rass, 1949; Sawyer, 1967;
Tokuya and Amaoka, 1980
2 Andriyashev, 1954; Barsukov, 1959;
Breder, 1941; Kobayashi, 1961a;
Marliave, 1975; Rass, 1949
1 Kobayashi, 1961b; Richardson and
Denhart, 1975
I Chapman and Townsend, 1938; Haryu
and Nishiyama, 1981
ous or large (Blenniini and Omobranchini) or completely lacking
(Nemophini). Teeth develop early in most species; these become
large (Nemophini) or hooked (Salariini) (Fig. 302E). Cirri may
develop at the end of the larval period. Members of the Salariini
have elongate pectoral fins (Fig. 302F).
Larvae of the northern families have an elongate body shape.
but they range from moderately elongate (Zaproridae) to ex-
tremely long and thin (Ptilichthyidae) (Fig. 303). Heads are
small, and initially pointed or rounded but become more point-
ed with development. Most species have a short to moderate
snout. Preanal length is highly variable. Generally, preanal length
is at least 50% BL, but it ranges from short (<50% BL in pre-
MATARESE ET AL.: BLENNIOIDEA
567
Table 144. Summary of Egg Characters in Blennioidea. Blanks indicate data are unavailable.
Egg Single
lype' or mass
Egg diameter
(mm)
Number of
oil globules
Altachmenl
processes or
ornamentation
Pigmentation
Incubation
(days)
Primary' sources
Blennioidea
Dactyloscopidae
Xenocephalidae
Notograptidae
Tripterygiidae
Clinidae
D
D Mass
D^
0.90-1.40
0.96-1.7
Few to
numerous
Several
Filaments at
one pole or
everywhere
Filaments
Embryo, yolk
with "pig-
ment
spheres'"
Chaenopsidae
D
Labnsomidae
D
Mass
1.15-1.33
1-6
Attach to each
other,
strands
Embryo, yolk
Blenniidae
D
Mass
0.58-1.6
X
0.4-0.96
0-several
Adhesive disk
or pedestal
Embryo, yolk
Bathymasteridae
Stichaeidae
D
D
Mass
Mass
0.99-1.1
1.37-2.5
Non-adhesive
mass
Adhesive
Cryptacanthodidae
Pholidae
Anarhichadidae
P^ilichthyidae
Zaproridae
Scytalinidae
D
D
D
Mass
Loose or
clumps
1.8
1.4-3.0
4.0-8.0
Adhesive
Dawson, 1982
16-22 Graham, 1939; Miller
et al., 1979; Ruck,
1973a; Ruck. 1980;
Shiogaki and Dotsu,
1973
12-40 Bamhart, 1932; Pa-
doa, 1956h; Shiogaki
and Dotsu, 1972b;
Sparta, 1948; Ste-
vens, unpubl.
Stephens et al., 1966
10 Breder, 1939
6-6 P Cipria, 1934, 1936;
Dotsu, 1982; Dotsu
and Moriuchi,
1980; Dotsu and
Oota, 1973; Dutt
and Rao, 1960; Eg-
gert, 1932; Fishel-
son 1963, 1976;
Fritzsche, 1978;
Hildebrand and Ca-
ble, 1938; Lebour,
1927; Mito, 1954;
Munro, 1955; Peters,
1981; Qasim, 1956;
Rao, 1970; Stevens
and Moser, 1982;
Thomson and Ben-
nett, 1953; Watson,
unpubl.; Wickler,
1957
13-15 Breder and Rosen,
1966; Fitch and
Lavenberg, 1975;
Matarese, unpubl.
2 1 Breder and Rosen,
1966; Hart, 1973;
Marliave, 1975;
Matarese, unpubl.;
Peppar, 1965; Shio-
gaki, 1983; Wourms
and Evans, 1974
Hart, 1973
42-70 Breder and Rosen,
1966; Marliave,
1975; Matarese, un-
publ.; Sawyer, 1967
Barsukov, 1959; Bred-
er and Rosen, 1966;
Matarese, unpubl.
I D = demersal.
^ Ophichni-ovoviparous George and Spnnger (1980). Clmmi-viviparous (Pennth, 1969: Hoese. 1976).
'Usually 7-14 days.
568
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 145. Summary of Larval Size at Selected Developmental Stages in Blennioidea (mm SL). Blanks indicate data are unavailable.
Taxon
Hatching
Notochord flexion
Special prejuvenile
Juvenile
Blennioidea
Dactyloscopidae
Xenocephalidae
Notograptidae
Tripterygiidae
2.7-6.1
4.8-<9.4
None
>11.0
Clinidae
5.5-6.7
by 11.4-14.3
None
> 16-25
Chaenopsidae
>16-17
Labrisomidae
4.1
4.9-6.9
None
> 19-25
Blenniidae
2.0-5.4
3.6-10.75
4.6-26.0 Salariini
6.4-26.0
Bathymasteridae
5.5-6.0
<10.0
None
ca. 40.0
Stichaeidae
6.5-12.5 (8-9.5)
12.0-20.0(13-15)
None
>25.0
Cryptacanthodidae
10.0-11.0
<18.0
Neustonic
ca. 30.0
Pholidae
10.0-12.5
ca. 19.0-30.0
None
>30.0
Anarhichadidae
17.0-18.0
<20.0
None
ca. 40.0
Ptilichthyidae
<20.0
None
ca. 114.0
Zaproridae
<12.0
<17.0
Neustonic
Scytalinidae
flexion Stichaeidae) to very long (Pholidae) (Fig. 303B. D). The
family Anarhichadidae includes one genus with a long, thin
bodied larva (Anarrhtchthys) and another with only a moder-
ately elongate larva (Anarhichas) (Fig. 303E). The monotypic
family Ptilichthyidae has a unique larval form — it is highly elon-
gate with a small head and extended postanal body (Fig. 303n.
Pigmentation characters. — Pigmentation is typically sparse for
most families within this infraorder, and tends to be added
subcutaneously with development. However, four families [Za-
proridae, Anarhichadidae, Cryptacanthodidae, and some Nem-
ophini (Blenniidae)] have larvae with dense body pigment that
is not typical of the Blennioidea (Figs. 303C, E, and G). Im-
portant pigment areas are along the ventral body midline and
in the gut area (Table 146).
Head. —Eyes are pigmented prior to hatching in all known groups.
Pigment is generally absent or light during notochord flexion
but usually increases dorsally, over the brain, by the time flexion
occurs. Additionally, postflexion larvae may have pigment on
the snout, mouth, and in the opercular area.
Gut. — Preflexion larvae in most species have peritoneal and
some dorsolateral pigment. In families that have a gas bladder
(e.g., Tripterygiidae, Labrisomidae, and Blenniidae), pigment is
present on its dorsal surface (Fig. 302A, D). Ventral pigment
may or may not be present. During notochord flexion, pigment
increases on the lateral gut surface, and becomes subcutaneous
in postflexion larvae.
Trunk and tail. — This is the most important pigment area for
Table 146. Summary of Some Pigmentation Characters in Larvae of Blennioidea. Key: D, dorsal; A, anal; P, pectoral; V, ventral; C,
caudal; +, present; O, absent; An, anterior; Po, posterior; \. increasing; j, decreasing; -, with development; and O - +, unpigmented initially,
becoming pigmented with development.
Head
Gut
Taxon
Eye at
hatch-
ing
Brain
Jaw
Snout
Oper-
cle
Isthmus
Nape
Antenor
Doi^al
Ventral
Lateral
Tripterygiidae
+
+
O
o
O
o
O
o
+
O, +Po
o
Clinidae
+
o
o
o
O
o
O - +
o
+
+ Po
o
Chaenopsidae
o
o
o
O
o
O
o
o
O
o
Labrisomidae
o
o
o
o
o
o
o
+
+
o
Blenniidae
+
o, -h-T
o
o
o
o
o
o
+ T
0-+1
+ -I
Bathymasteridae
+
O- -1-
o
o
o
+
o
o
+
+
o
Stichaeidae
+
o, +
o, +
o
o
-1-
o
o
+ -i
-1-
o
Cryptacanthodidae
-1-
+ T
+
+
+ [
o- -^
+
o
+ -1
o
o
Pholidae
-1-
o-i
o
o
o
+
o
o
+
-1-
o
Anarhichadidae
-1-
+ 1
-t-
+
+ 1
-1-
o - +
o
+ T
o- +
+ -!
Ptilichthyidae
-1-
+
o
o
+
o
-1-
+
+
O - +An
Zaproridae
-1-
+ T
+
+
+ 1
4-
-1-
o
+
o
+ 1
MATARESE ET AL.: BLENNIOIDEA
569
identifying specific groups within this infraorder. Except for the
densely pigmented famihes hsted above, pigment along the dor-
sal body midline is rare in preflexion larvae. With development,
pigment may increase along the dorsal midline or on the nape.
Initially, lateral pigment is either absent or consists of a few
spots internally along the notochord. After notochord flexion,
internal and external pigment can increase ventrolaterally. or
above and below the notochord (Stichaeidae, Bathymastendae,
and Pholidae). Typically, a series of ventral midline melano-
phores occurs in preflexion larvae. Although these melano-
phores may be absent in some families (Chaenopsidae, some
Tripterygiidae), a number of families have larvae with up to 50
melanophores here (e.g.. Blenniidae). The number, size, and
shape of these melanophores can be very important when iden-
tifying groups. These spots may change shape with development
(becoming y-shaped in Tripterygiidae and some Blenniidae),
decrease in number (some Blenniidae and Stichaeidae), or be-
come subcutaneous (Stichaeidae).
Fins. — With the exception of zaprondsand some blenniids, fins
are rarely pigmented in preflexion larvae. After notochord flex-
ion pigment develops on the various fins of blenniids, anarhich-
adids, and ptilichthyids (Table 146).
Hypural margin. — Pigment in the caudal area is usually lacking
in preflexion larvae, and in postflexion larvae its presence is
limited to a few families (Table 146).
Meristic characters.— The number of dorsal fins varies from one
to three and in most families some combination of spines and
rays is present, with spines predominating. Tripterygiids, clin-
ids, and labnsomids may have up to three dorsal fins, the first
two composed of spines. The total number of dorsal elements
is highly variable but in some groups (stichaeids, anarhichadids,
and ptilichthyids) well over 100 elements are present. The anal
fin in most groups may include 1-2 spines. Stichaeids may have
up to 5 anal spines. Information on the caudal fin is incomplete.
In addition, from data available in the literature, principal rays
and branched rays are not consistently distinguished. Most groups
have between 9 and 15 (usually about 12-13) principal caudal
fin rays and about 25-30 total caudal fin rays. All possess a
pectoral fin with as few as 3 (labrisomids and clinids) or as many
as 25 (zaprorids) fin rays. Pelvic fins can be present or absent.
The northern families, except some stichaeids and pholids, lack
pelvic fins. Tropical families usually possess thoracic pelvic fins
with 1 spine and fewer than 5 rays (mostly 2-3 soft rays).
Vertebral counts are unknown for many blennioids or are
based on few specimens. The number of vertebrae is highly
variable within some families (e.g., stichaeids, blenniids, an-
archichadids). In general, tropical families have a lower verte-
bral count than do northern families.
The order of fin ray development is highly variable in the
Blennioidea. Information available on this is also inadequate,
since in most studies reviewed here larvae have not been cleared
and stained to determine the onset of ossification. In the tropical
families where notochord flexion occurs as early as 3.6 mm, fin
ray development may begin as early as 2.5 mm. Caudal fin rays
develop first in clinid and labrisomid larvae, followed by the
remaining fin rays soon after notochord flexion is complete.
Typically, pectoral fin rays develop first in blenniid larvae (Blen-
niini and Salariini). In Ombranchini larvae (Blenniidae), the
pectoral fin rays and caudal fin rays develop simultaneously.
Among the northern families, data are insufficient to allow any
generalizations. Fin rays begin forming at 9-1 5 mm in stichaeid
larvae (usually caudal fin rays first) but may not be complete
until larvae are 30 mm (Fig. 303B). Zaprorid and cryptacan-
thodid larvae begin caudal ray development about the time
notochord flexion occurs. Fin ray development in ptilichthyid
larvae begins with the dorsals and second anal at 40 mm.
Relationships
Although the scope of the available egg and larval data within
the Blennioidea is limited, early life history characters reviewed
here do not support the cohesiveness of this group. Due to a
lack of unifying characters, the infraorder Blennioidea, as pres-
ently arranged, probably does not form a monophyletic group.
Early life history characters appear to be more useful in clarifying
relationships between families or within families rather than
Table 146. E.xtended.
Trunk
Internal
Dorsal
Dorso-
Medio-
Ventro-
Ventral
noto-
Hypural
margin
lateral
lateral
lateral
margtn
chord
margin
Fin base
Diagnostic
+ Po
o- +
O
o
-1-
o
o
P
Anus, ventral midline
+ -T
O- -1-
+ 1
o
-t-
+
o
O
Gut, ventral midline
o
o
o
o
o
o
o
O
Lack of pigment
O, +Po
o
o
o
+
o
o
O
Swimbladder, ventral
midline
O - + An
o - +
O- +1
o-o, +
+ -1
-1-
o -o, +
o
- PV
Gut. ventral midline
O- +
o
o
O- -H
-t-
-1-
O, -t-
o
Urostyle or lateral
cross-checking
o- +
o
+ -T
O- H-
+ -1
+
o - o, +
o
Gut, anus, ventral
midline
+ T
+ T
-t-
+ T
-1-
-
o
o
Dense body
O-O, -1-
o
+
o
-1-
+
0--0, +
o
Gut, ventral midline
+ T
+ 1
+ 1
+ 1
-1-
+
o
o
- DA
Dense body, fins
+
o
o
o
+ T
o
o
APO
-C
Gut, dorsal and ventral
margin, caudal fin
+ I
+ 1
+ 1
+ 1
+ 1
-
+ T
AP
-TD
Dense body
Fig. 302. Blennioidea larvae (tropical forms): (A) Enneaplerygius atriceps (Tripterygiidae), 5.8 mm (from Miller et al., 1979 described as
Triplerygion atriceps); (B-C) Heterostichus rostralus (Clinidae), 6.5 mm, 21.2 mm; (D) Parachnus integripmnis (Labrisomidae), 7.2 mm; (E-F)
Istiblennius zebra (Blenniidae), .1.3 mm, 1 1.0 mm (from Miller et al., 1979); (G) Enche/vunis bnmneolus (Blenniidae), 3.2 mm (from Miller et
al., 1979).
-^r.^^^rrr^r';^^-7r^^^P777:^^^c^^^,^^,^^.^
Fig. 303. Blennioidea larvae (northern forms): (A) /?OAi(?M//M.S7orrfaM/(Bathymasteridae), 10.4 mm; (B) -4«op/arf/H«pwrpwr«few.y(Stichaeidae),
12.0 mm; (C) Lyconectcs alcutensis (Cryplacanlhodidac), 16.0 mm; (D) Pholis sp. (Pholidae), 23.0 mm; (E) Anarhichas lupus lupus (Anarhi-
chadidae). 24.5 mm (from Barsukov, 1959); (F) Pnlichlhys goodei (Ptilichlhyidae), 24.7 mm (from Richardson and Dehart, 1975); (G) Zaprora
i(/f «Mi (Zaproridae), 16.0 mm (from Haryu and Nishiyama, 1981).
572
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 147. Summary of Selected MERisTrcs in Blennioidea. Blanks indicate data are unavailable.
Fins
Dorsal
Anal
Pectoral
Pelvic
Tax on
Spines
Rays
Spines
Rays
Blennioidea
Dactyloscopidae
0-VII + V-XVIII
14-36
22-41
12-16
1,3
Xenocephalidae
Notograptidae
Tripterygiidae
Clinidae
III-VII + X-XXIV
III + XXIV-LXXXIV
7 - 10 21
7-15 O-II 14-30 10-19
1-14 II 14-62 3-18
5
1
1,2-3
1,2-3
Chaenopsidae
XVII-XXXIII
10-34
18-38 12-14
1,3
Labrisomidae
III + I-IV + XX-LII
7-14 I-Il
15-28
3-17
O-I, 0-3
Blenniidae
III-XVII
9-119 II
10-119 10-lJ
1, 2-4
Bathymasteridae
Stichaeidae
II
XXII-CXXI
39-49
0-43
I-II
I-V
27-36
24-95
17-21
8-21
1,5
Absent or I, 1-5
Cryptacanthodidae
Pholidae
Anarhichadidae
LX-LXXVII
LXXIV-CCL
LXX-CCL
-
II
I-II
O-I
45-50
29-53
42-233
11-15
10-16
18-23
Absent
Absent or I, 0-1
Absent
Ptilichthyidae
Zaproridae
LXXXIII-XC
LIV-LVII
115-148
—
179-196
24-30
13
20-25
Absent
Absent
Scytalinidae
41-51
-
41-51
8
Absent
between Blennioidei infraorders, e.g., the similarity between
labrisomid and clinid larvae and the differences between larvae
in the various blenniid tribes.
Many of the families in Blennioidea include a large number
of intertidal forms and many of the similarities (e.g.. demersal
eggs, parental care, and advanced state of newly hatched larvae)
may be related to environmental conditions rather than to a
close phylogenelic relationship. Additional study on the com-
plete life history of these fishes is needed to identify unifying
characters, if any exist. Studies at the family level will improve
our knowledge of this unsatisfactorily defined group and facil-
itate outgroup comparisons.
(A. CM.) National Marine Fisheries Service, Northwest
AN15 Alaska Fisheries Center, 2725 Montlake Boule-
vard East, Seattle, Washington 981 12; (W.W.) Marine
Ecological Consultants, 531 Encinitas Boulevard,
Suite 110. Encinitas. California 92024; (E.G.S.) Na-
tional Marine Fisheries Service. Solithwest Fisheries
Center, PO Box 271, La Jolla, California 92038.
MATARESE ET AL.: BLENNIOIDEA
Table 147. Extended.
573
Pnncipal caudal
Caudal
Total
Primary sources
9-11
11-14
25-41
13-15
10-13
20-30
10-15
13-35
25-63
13, 19-23
10-14
10-15
10-14
9-16
20-33
19-119
40-50
39-57
28-135
14
14-16
35-39
49-54
3-8 + 3-9
14-43
29-72
46-113
13-15, 14
24-27
47-51
72-78
6-7 + 6-7
80-107
7-8 + 19-26
24-39
46-214
72-250
53-59
170-181
227-240
30-31
24-26
61-62
Bohlke and Caldwell, 1961; Bohlke and Chaplin, 1968;
Dawson. 1974a. 1975, 1976, 1982; Kanazawa. 1952;
Miller and Briggs. 1962; Myers and Wade. 1946
Munro, 1967; Nelson, 1976
Nelson, 1976
Bath, 1973; Leis and Rennis, 1983; Rosenblatt, 1959,
1960; Wheeler and Dunn. 1975
Bohlke, 1960b; George and Springer, 1980; Hoese, 1976;
C. Hubbs, 1952, 1953a; Pennth. 1969; Scott, 1955.
1962, 1966, 1967; Shen. 1 97 1 ; Springer, 1955, 1970;
Stevens and Springer, 1974
Bohlke, 1957; Greenfield, 1972; Johnson and Green-
field. 1976; Robins and Randall. 1965; Rosenblatt
and Stevens. 1978; Smith-Vaniz and Palacio. 1974;
Stephens, 1963, 1970; Stephens et al.. 1966
Bohlke and Robins. 1974; Bohlke and Sponger. 1961,
1975; C. Hubbs. 1952, 1953b; Rosenblatt and Parr,
1969; Rosenblatt and Taylor. 1971; Smith, 1957;
Sponger, 1954, 1955, 1958, 1959; Springer and Go-
mon, 1975b
Bath. 1976, 1978; Smith-Vaniz, 1975, 1976; Smith-
Vaniz and Springer. 1971; Springer, 1967, 1968,
1971, 1972a, 1972b, 1976; Springer and Gomon,
1975a; Springer and Smith-Vaniz, 1972; Springer and
Spreitzer, 1978; Stephens. 1970
NWAFC, unpubl.
Makushok, 1958; NWAFC. unpubl.; Shiogaki, 1980,
1981
NWAFC, unpubl.; Shiogaki, 1982
Makushok. 1958; NWAFC. unpubl.
Barsukov, 1959; Makushok, 1958; NWAFC. unpubl.
Makushok. 1958
Chapman and Townsend, 1938; NWAFC, unpubl.
NWAFC, unpubl.
Ammodytoidei: Development and Relationships
E. G. Stevens, A. C. Matarese and W. Watson
THE suborder Ammodytoidei consists of one family. Am-
modytidae, with 5 genera and about 1 8 species. These are
small (less than 100-350 mm SL), elongate fish occurring in the
littoral and neritic waters of the Atlantic, Indian, Pacific, and
Arctic Oceans. Adults form schools but also bury themselves
in the sand. They are commercially valuable in the North Sea
and off Japan.
The systematic position of Ammodytoidei. reviewed by
Duncker and Mohr ( 1 939). is unresolved, although the suborder
is considered a perciform derivative by Berg ( 1 940), Greenwood
et al. ( 1 966), and Gosline ( 1 97 1 ). A second family. Hypoptych-
idae. has been included in this suborder by these authors and
by Robins and Bohlke ( 1 970), but was removed to the suborder
Gasterosteoidei by Ida (1976), who considered it a preperco-
morph family on the basis of jaw and caudal osteology, egg size,
and reproductive behavior (see Fritzsche, this volume).
Development
The ammodytid genera, Gymnammodytes, Hyperoplus, and
Ammodytes (11 species) are temperate and boreal; Bleekeria
and Embolichthys (7 species) are more tropical in distribution.
The confused nomenclature of the North Atlantic species was
clarified by the synonomies in Reay (1970) and Russell (1976),
where summaries of early life history data were also given. Other
larval descriptions were given by Fage (1918) for Gymnam-
modytes; by Altukhov (1978), Kobayashi (1961c), Norcross et
al. (1961). Richards (1965). Scott (1972). and Senta (1965) for
Ammodytes; and by Macer (1967) for North Atlantic species.
To date, eggs of 6 species and larvae of 9 species of these genera
have been described. No early life history data are available for
the tropical genera.
Eggs
Eggs of the six species that have been described are demersal
and adhesive, forming clumps on sandy substrates in shallow
water. Eggs, probably loosened by tidal currents, have been
collected in plankton nets (Williams et al. 1964; Senta, 1965).
Russell (1976) summarized studies made on eggs resulting from
artificial fertilization. Incubation time ranges from 2.0 to 12.5
weeks. Eggs are irregularly shaped, but generally spherical, from
0.67 to 1.23 mm in diameter, with a single yellow oil globule.
0.17 to 0.42 mm. Embryos develop specific dorsal and ventral
pigment, pigmented eyes, a moderate finfold. and pectoral buds
prior to hatching at about 3.6 mm.
Larvae
Morphology. — harvdiC of Ammodytidae typically are elongate,
with rounded snouts which become pointed with age, and pre-
anal length slightly more than 50% body length (Fig. 304). Newly
hatched larvae range from 3.0 to 4.6 mm body length. In newly
hatched and preflexion larvae the anus does not extend to the
edge of the moderately wide finfold but opens to the side. No-
tochord flexion occurs at 10 to 12 mm body length in most
=====:----r^/f0<^^^
\\\W'
■* » •■i,...jj-^..^.;
:i^>;^i^;<g.^^
Fig. 304. Larvae of: (upper) Hyperoplus lanceolalus. 16 mm, redrawn by H. Orr from Einarsson (1951); (middle) Ammodytes hexaplerus. 16
mm; and (lower) Ammodytes marimis, 16 mm. redrawn by H. Orr from Einarsson (1951).
574
STEVENS ET AL.: AMMODYTOIDEI
575
Table 148. Selected Pigment Characters of Ammodytidae Larvae. 0 = absent, + = present,
with development, po = posterior, an = anterior.
increasing with development, i = decreasing
Species
Stage
Body length
(mm) Jaws
Snout Brain
Nape
Dorsal
midline
Fin-
Caudal fold
Gymnammodyles semisquamaliis prefiexion
flexion
postflexion
prefiexion
postflexion
flexion
postflexion
G. cicerellus
Hyperoplus lanceolatus
H. immaculatus
Ammodyles tobianus
A. marinus
dubius
americanus
A. hexapterus
preflexion
flexion
postflexion
preflexion
flexion
postflexion
preflexion
flexion
postflexion
preflexion
flexion
postflexion
4.8
7.0
11.8-38.0
5.5-9.0
13.0
26.0
4.0-5.0
7.5-12.0
16.0-27.0
4.5-6.0
7.5-11.0
19.0-33.0
7.0-8.0
11.0-13.0
16.0-31.0
7.0 0
11.0-25.0 0
6.0 0
11.0-25.0 0, +
0
0
+
0
0
0
0
0
0
0
0
OT
0
0
OT
0
+ ,0
+
0
0
+ .0
0
+
0 +
0, + +
0 01
0 +
+ +
0 0
0 +
0 +,0
0
+
+
0
0
01
+
+1
0
0
+
+1
an
an
an
+ an
+ ,0 an
+ ,0 0
an
an
+
an
an
an
0
near tail
T +
po 1/3
+
po 1/4
0
near tail
I
0
po 1/4
I
0
0
01
0
+
+
+
+
0
I
0
+
+
0
0
01
+
+
+
0
+
+
Cameron, 1959
Macer, 1967
0 Page, 1918
Einarsson, 1951
Macer, 1967
Macer, 1967
Einarsson, 1955
Macer, 1967
Einarsson. 1951
Macer, 1967
Kobayashi, 1961c
NWAFC, unpubl.
species, and transformation to juveniles occurs at about 40 mm.
The caudal fin is the first to ossify, followed by the pectorals,
then the dorsal and anal. The median fin rays form in the pos-
terior part of the body, and ossification proceeds forward. Dur-
ing larval development the body thickens somewhat, but main-
tains its elongate shape. All adult Ammodytidae have protrusible
upper jaws, but Gymnammodytes semisquamatus is the only
species in which this character is reported in larvae as small as
9 mm (Cameron, 1959). Postflexion larvae oi Hyperoplus de-
velop vomerine teeth which persist in the adult, while Gym-
nammodytes postflexion larvae develop both vomerine and pre-
maxillary teeth which disappear at about transformation. During
the larval period Gymnammodytes. Hyperoplus. and .Ammo-
dyles are pelagic. Juveniles and adults are both pelagic and
benthic.
Pigment. — Pigment can be a useful diagnostic feature among
the larvae of Ammodytidae, especially the location and devel-
opment of melanophores on the ventral gut margin, the dorsal
body margin, and the caudal area, i.e., the tip of the notochord
and the edge of hypural elements. These pigment characters are
summarized in Table 148. All species have a row of melano-
phores dorsally on the gut, beginning at or just posterior to the
cleithrum, and a postanal row on the ventral body margin from
the anus to the tail. The dorsal gut pigment becomes obscured
with growth. Specific variations in pigment patterns can be seen
in the 16 mm specimens illustrated in Fig. 304. At this length,
dorsal midline pigment forms a complete row in H. lanceolatus,
but occurs only on the posterior quarter in .-i. hexapterus and
.4. marinus: and ventral gut pigment extends the length of the
gut in H. lanceolatus and .4. marinus but is found only on the
anterior ventral gut of .4. hexapterus. Pigment patterns of .1.
marinus, .4. dubius and .4. americanus are nearly identical (Ma-
cer, 1976) although Richards (1982) has noted diflTerences in the
ranges of melanophore numbers, especially on the anterior ven-
tral gut (stomach) and dorsal midline (supradorsal). Pigment
appears variously on the head, increasing with age in all species
reported. The only reported decrease in pigmentation is on the
dorsal and ventral margins of G. cicerellus (Page, 1918). G.
semisquamatus has pigment on the ventral finfold margin, the
only ammodytid species for which finfold pigment has been
noted (Cameron, 1959).
Aferistics. — Fin ray and vertebral counts for the family Am-
modytidae are: Vert 54-78; D 40-69; A 14-36; Pec. 10-16; Pel.
0-1,5; and C 15-17 prin., 1 3 branched. In all genera the number
of precaudal vertebrae exceeds the number of caudal. Robins
and Bohike (1970) report 9 + 8 principal caudal rays for Em-
bolichthys sarissa. but all other ammodytid species, including
E. mitsukurit, have 8 + 7. Embolichthys is the only genus with
pelvic fins, which are thoracic. The caudal fin is the first to form,
followed by the pectorals, dorsal, and anal. Posterior rays of the
median fins form first and development proceeds forward. Fin
formation is completed by 30 to 40 mm body length.
Relationships
Although early life history data of the suborder Ammody-
toidei do little to clarify its phylogenetic position, larval pigment
patterns and myomere- vertebrae counts are useful in separating
sympatric species (Macer, 1967). General characters, such as
the well developed state of newly hatched larvae and the se-
quences of fin development, are shared with other perciform
derivatives and relatives, but essentially the problem of the
systematic position of ammodytids is not yet resolved.
(A. CM.) National Marine Fisheries Service, Northwest
AND Alaska Fisheries Center, 2725 Montlake Boule-
vard East, Seattle, Washington 98112; (E.G.S.) Na-
tional Marine Fisheries Service, Southwest Fisheries
Center, PO Box 27 1 . La Jolla, California 92038; (W.W.)
Marine Ecological Consultants, 531 Encinitas Boule-
vard, Slute 1 10, Encinitas. California 92024.
Icosteoidei: Development and Relationships
A. C. Matarese, E. G. Stevens and W. Watson
THE suborder, Icosteoidei, consists of one family, Icosteidae,
with a monotypic species Icosteus aenigmaticus (Nelson,
1976). Adults inhabit the epi- and bathypelagic areas of the
North Pacific Ocean from southern California to Japan. Adults
may exceed 2 m in length and have dark flaccid bodies; a char-
acteristic implied in the common name of the species, raglish
(Hart, 1973).
The systematic position of this group and its designation as
an order or suborder is not well established. Greenwood et al.
( 1 966) consider it a suborder of Perciformes while Gosline (1971)
elevates it to an order, Icosteiformes, a probable perciform de-
rivative.
Development
The only early life history data previously published is a brief
description of the egg (Allen, 1968). Icosteus aenigmaticus eggs
are commonly collected in ichthyoplankton surveys off the Pa-
cific coast of North America [California Cooperative Oceanic
Fisheries Investigations (CalCOFI) and Northwest and Alaska
Fisheries Center (NWAFC)], but larvae (mostly preflexion) are
infrequently found and a complete size series from hatching to
transformation is not presently available. Larvae may move
offshore or into deeper waters. The first published description
and illustration of the larvae from pre- to postflexion stages are
provided here, based on National Marine Fisheries Service
(NMFS) collections. Although /. aenigmaticus ]uven\\e^ undergo
a marked transformation to the adult stage, little information
is available concerning this change (Hart, 1973).
Eggs
The pelagic egg of /. aenigmaticus ranges in diameter from
2.8 to 3. 1 mm (Fig. 305). A large, sometimes irregular, oil glob-
ule with a diameter of 0.42 to 0.60 mm is present. The oil globule
usually decreases in size with development. The chorion is
smooth, sometimes amber or rose colored. Early stage egg yolks
are frequently opaque, although later stages have a clear, un-
segmented yolk. During the middle stage of development, em-
bryos have pigment along the dorsal body as well as on the yolk
and oil globule. Late stage embryos have functional mouths,
pectoral buds, and very wide finfolds. Scattered pigment occurs
on the eyes, snout, jaws, and dorsal head. The dorsal surface of
the gut is pigmented. Along the dorsal and anal finfolds, three
or four clusters of melanophores appear at each distal edge.
Melanophores also appear above and below the tail in the caudal
finfold. An irregular double row of melanophores extends the
length of the dorsal body margin. A few mediolateral spots
appear anteriorly. Occasionally, pigment occurs along the ven-
tral body margin.
Larvae
Morphology.— Nevj\y hatched larvae of/, aenigmaticus are 6.5
mm NL; yolk material may persist until larvae are 10 mm.
Flexion begins at about 1 1 mm and is complete at about 1 7
mm SL. The size at transformation is not known, but fin de-
velopment is almost complete by 28 mm. The body, surrounded
by a wide finfold, is very soft. Preflexion larvae have small heads
with rounded snouts and long tapering bodies (Fig. 306). Dor-
sal and ventral finfolds are wider than the body. During flexion
the body thickens and becomes more robust, especially ante-
riorly. Postflexion larvae have a robust head and gut and a
tapering trunk (Fig. 306). Preanal length is less than 50% body
length. A series of preopercular spines appears during late flex-
Pigment. — New\y hatched larvae of /. aenigmaticus display es-
sentially the same eye, head, gut, body, and finfold pigment as
the embryos. With increasing size the head and gut usually
become increasingly covered with discrete spots. Dorsal body
margin pigment is present throughout larval development, while
the amount of lateral and ventral body margin pigment varies
and is relatively sparse. The characteristic embryonic caudal
pigment persists in the developing larvae, becoming less prom-
inent but remaining as scattered melanophores on the hypural
margin and fin ray bases. In general, postflexion larvae are less
pigmented except on the head. Pelvic and pectoral fin bases and
pelvic rays acquire melanophores during postflexion.
Meristics. — Icosteus aenigmaticus larvae have the following ver-
tebral and fin ray counts: Vert. 66-68; D 55; A 39; Pec. 21; Pel.
1,4; and C 9 -(- 8 = 17 (NWAFC files). These counts conform
Fig. 305. Egg of Icosteus aenigmaticus: 2.8 mm, drawn by H. Orr.
576
MATARESE ET AL.: ICOSTEOIDEI
577
Fig. 306. Larvae of Icosteus aenigrnaticus from top to bottom: 9.5 mm SL; 10.2 mm; and 28.5 mm SL, drawn by H. Orr.
to those for the adults except adults lack a pelvic fin (Abe, 1954;
Miller and Lea, 1972; Hart, 1973). The caudal fin contains the
perciform number of principal rays, 1 7, with 6-9 procurrent
rays on each side. Pectoral fin blades are present at hatching
and rays form during flexion. Pelvic fin rays begin development
during flexion and are complete in postflexion larvae. At what
size the pelvic fins disappear is not known. The last fin rays to
form are the dorsal and anal, with their anlagen appearing in
the middle of the posterior half of the finfolds at about mid-
flexion. Formation of these fins proceeds forward and toward
the body margin (Fig. 306). The largest larva available, 28.5
mm, has the complete fin ray complement.
Relationships
The foregoing brief description of the eggs and larvae of /.
aenigrnaticus provides some additional information toward the
understanding of the life history of this unique but poorly under-
stood fish. Characters discussed here (e.g., sequence of fin for-
mation and meristics) help support its position among perciform
relatives. Sequence of fin formation and reduced number of
pelvic fin rays are blennioid-like characters, and 1 7 principal
caudal fin rays are the typical percoid number. Eggs, larvae, and
early juveniles superficially resemble stromateoid fishes but ad-
ditional data are needed before a precise relationship can be
determined. To understand this family more fully, we need
information regarding the critical juvenile phase as well as a
complete osteological examination from preflexion larvae to
adults.
(A. CM.) National Marine Fisheries Service, Northwe.st
AND Alaska Fisheries Center, 2725 Montlake Boule-
vard East, Seattle, Washington 98112; (E.G.S.) Na-
tional Marine Fisheries Service, Southwest Fisheries
Center, PO Box 27 1 , La Jolla, California 92038; (W.W.)
Marine Ecological Consultants, 53 1 Encinitas Boule-
vard, Suite 1 10, Encinitas, California 92024.
Zoarcidae: Development and Relationships
M. E. Anderson
THE eelpouts, Zoarcidae, comprise a monophyletic group of
about 200 valid species of marine fishes in 44 genera (Table
149; Anderson, 1984). About 20 additional undescribed species
are known to me from collections around the world. Most zoar-
cids live on the bottom in deep water in boreal seas, but 1 1 are
known from intertidal areas, especially in temperate South
America. Twenty-two species are known from tropical-sub-
tropical areas and all of them live in deep water (640-5,320 m).
Fourteen species are known from both shallow and deep waters
of Antarctica and subantarctic regions. Two deep-living genera,
Lycodapus and Melanostignia, are coastal or thalassobathyal,
deep-pelagic forms that seem to occur in greatest numbers where
their zooplankton prey concentrate (Belman and Anderson, 1 979;
Anderson, 1981). Thus the family is stenothermic and adapted
to very low temperatures (mostly below about 8° C).
Table 149. Distribution, Ecology and Selected Meristics of Zoarcidae.
No. ol
species
Distribution
Genus
Ecology
D
Aiakas
1
SW Atlantic
Benthic; slope
88-89
Andriashevia
1
NW Pacific
Benthic; slope
144
Austrolycichthys
4
E trop. Pacific; W trop.
Atlantic; Antarctica
Benthic; slope
O-ll, 87-104
Bilahria
1
NW Pacific
Benthic; shelf
110
Bothrocara
8-9
NW Pacific to Peru
Benthic; slope
108-125
Bothrocarina
2
NW Pacific
Benthic; shelf-slope
1, 106
Cwssoslomus
2
SW Atlantic
Benthic; shelf
96-108
Dadyanos
I
SW Atlantic
Benthic; shelf
104-116
Davidijordania
4
NW Pacific
Benthic; shelf
85-118
Derepodichthys
1
NE Pacific
Benthic; slope
110-116
Exechodontes
1
W trop. Atlantic
Benthic; slope
80-86
Gymnelopsis
4
NW Pacific
Benthic; shelf-slope
I, 79-112
Gymnelus
5
N Pacific, Arctic
Benthic; shelf
I, 76-101
Hadropareia
1
NW Pacific
Benthic; shelf
106-113
Hadwpogonichthys
1
NW Pacific
Benthic; slope
126-128
Iluocoetes
2
SE Pacific-SW Atlantic
Benthic; shelf
84-98
Krusensterniella
4
NW Pacific
Benthic; shelf
XLV-LVII, II-XXVI
37-64
Lycenchelys
40-41
Worldwide, except Indo-Pac.
Benthic; slope-abyss
0-1,94-132
Lycodapus
13
NW Pacific to subantarctic
Deep pelagic
70-98
Lycodes
46-51
Boreal seas; South Africa
Benthic; shelf-abyss
85-120
Lycodichthys
2
Antarctica
Benthic; slope
l-II. 83-89
Lycodonus
4
N and S Atlantic
Benthic; slope
98-112
Lycogrammoides
NW Pacific
Pelagic ?
62
Lyconema
NE Pacific
Benthic; shelf-slope
100-107
Lycozoarces
NW Pacific
Benthic; shelf
II, 62-67
Macrozoarces
NW Atlantic
Benthic; shelf
92-103, XVI-XXIV,
16-30
Maynea
SW Atlantic-SE Pacific
Benthic; shelf
119-127
Melanosligma
7
Worldwide
Deep pelagic
76-95
Nalbantichthys
N Pacific
Benthic
slope
143-152
Notolycodes
SW Atlantic
Benthic
slope
85-88
Oidiphorus
SW Atlantic
Benthic
slope
56-61
Ophthalmolycus
3
Chile to Antarctica
Benthic
slope
O-I, 87-103
Pachycara
9
Worldwide
Benthic
slope-abyss
0-1,95-113
Phucocoetes
SW Atlantic-SE Pacific
Benthic
shelf
101-107
L
Piedrabuenia
SW Atlantic
Benthic
slope
108-113
Pogonolycus
SW Atlantic
Benthic
shelf
86-88
1
Puzanovia
NW Pacific
Benthic
slope
135-147
Taranetzella
N Pacific
Benthic
slope-abyss
84-89
Zoarces
3
NE Atlantic; NW Pacific
Benthic
shelf
72-94, 0-XIX, 14-
-27
Genus A.
Coast of California
Benthic
shelf-slope
97-107
Genus B.
Scotia Sea
Benthic
slope
_
Genus C.
Bering Sea
Benthic
slope
—
1
Genus D.
SW Atlantic
Benthic
slope
77-83
Genus E.
2
SW Atlantic
Benthic
slope
92-96
578
ANDERSON: ZOARCIDAE
579
Development
As far as known, almost all eelpouts are oviparous, laying
relatively few. large eggs. The exception is the genus Zoarces.
which is viviparous. There are three species of Zoarces, the
common European eelpout, Z. viviparus (Linnaeus), and two
little known, northwestern Pacific species, Z. gillii Jordan and
Starks and Z. ekmgatus Kner. Viviparity in the European eel-
pout has been known since the Middle Ages (Schonevelde. 1624).
but of the two Pacific species, females with embryos are known
only in Z. gillii (Anderson. 1984).
Among benthic, oviparous species, nest building with parental
guardianship is probably common. Nesting has been directly
observed in Macrozoarces americanns (Olsen and Merriman,
1946), Gymnelns viridis (Emery, 1973), Lycodes pacificus (Lev-
ings, 1969) and Phucocoetes latitans and Ihiocoetes effusus
(Gosztonyi, 1977). Probably most, if not all, the other South
American intertidal zoarcids discussed by Gosztonyi also build
and guard nest sites. Pelagic spawning occurs in Lycodapus and
Melanostigma. Markle and Wenner( 1 979) found Melanostigma
ailanticum may utilize the sea bottom as a concentration in-
terface for group spawning. Bottom trawl-caught ripe individ-
uals had parasite loads more typical of deep-demersal fishes in
the western North Atlantic. However, Anderson (1981) reported
Lycodapus mandihularis to have a parasite fauna similar to
other midwater fishes in Monterey Bay, California. Early ju-
veniles were caught in midwater at all depths inhabited by adults.
This suggests L. mandihularis does not shoal near the bottom
for spawning.
Eggs
Spawned zoarcid eggs have been described from field obser-
vations for only seven species (Table 1 50). Egg descriptions are
Table 149. Extended.
Fin rays
Venebrae
A
P
c
Precaudal
Caudal
Sources
67-69
18-19
10
26
65-66
This report; Gosztonyi (1977)
123
Absent
—
22
125
Fedorov and Neyelov (1978)
70-89
15-19
9-10
20-25
72-87
This report
93-94
15-16
10
23
95-96
This report; Lindberg and Krasyukova (1975)
92-109
13-17
0-13
18-24
93-108
This report
95
10-11
11
19
94
This report
68-78
16-17
10
28-32
67-76
This report; Gosztonyi (1977)
89-95
16-17
10
21-24
84-93
This report; Gosztonyi (1977)
68-90
12-17
10
20-23
77-97
This report; Lindberg and Krasyukova (1975)
94-101
10-11
8-9
22-26
92-98
Anderson and Hubbs (1981)
73-79
13-15
10
19-21
72-78
This report; DeWitt (1977)
73-97
9-12
5-8
16-23
73-95
Anderson (1982)
69-85
9-14
9-12
17-26
65-84
Anderson (1982)
86-92
13-15
7-8
24-28
83-89
This report
112-114
19
11
23-24
109-110
Fedorov ( 1 982)
65-82
15-19
7-9
19-24
62-79
This report; Gosztonyi (1977)
71-103
11-12
5-7
19-25
80-97
This report; Lindberg and Krasyukova (1975)
80-112
13-21
9-13
20-30
77-118
This report; Andriashev (1955b)
58-86
5-9
8-12
13-19
59-85
Peden and Anderson (1978, 1981)
67-92
14-24
10-12
19-26
65-104
This report
66-75
15-17
11
23-24
68-70
This report; DeWitt (1962a)
83-93
14-17
7-9
21-25
85-105
This repon
52
9
8^
14
53
This report
90-96
15-17
12
20-21
86-93
This report; Gotshall (1971)
49-54
13-15
13-15
15-17
50-55
This report; Toyoshima (1981)
103-125
17-20
9-10
25-28
105-118
This report
95-103
14-16
7
29-30
89-98
This report; Gosztonyi (1977)
62-80
6-9
8-10
18-23
62-81
This report; Parin (1977)
121-127
6
7-10
25
119-125
Schultz(1967)
69-72
18-21
11
23-26
66-69
This report; Gosztonyi (1977)
45-54
16-19
7-9
15-17
43-50
This report; Gosztonyi (1977)
69-87
14-18
10
22-23
72-88
This report
77-97
14-19
10-12
25-32
74-91
This report
78-85
14-16
10
24-27
75-83
This report; Gosztonyi (1977)
98-104
17-18
8
24-25
95-101
This report; Gosztonyi (1977)
72-74
17
9
20
71-73
This report; Gosztonyi (1977)
115-128
9-12
9-12
22-24
110-125
Fedorov (1975); Amaoka et al. (1977)
71-76
15
8
20
69-74
This report; Andriashev (1952)
64-90
16-21
9-11
21-26
80-106
This report; Schmidt (1917)
83-93
13-14
10
27-28
79-84
This report; Cailliet and Lea (1977)
—
—
—
—
—
Tomoet al. (1977)
—
—
—
—
—
Bond and Stein (in prep.)
64-69
17-19
9-10
19-21
66-73
This report
68-73
9-14
7
24-27
67-75
This report
580
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 150. Data on the Eggs and Larvae of Zoarcidae Known to Date.
Eggs
rv
Size
range, mm
Oil
globule
Incubation
(months)
Hatch size,
mm
Species
Descnption
Illustration
Sources
Gymnelus viridis
3.2-4.6*
_
_
-20-25
X
X
Rass(1949)
Macrozoarces americanus
-6.0-7.0
2.5-3.5
28-31
X
X
White (1939); Olsen
and Merriman
(1946)
Zoarces viviparus
2.8-3.2*
—
4
32-40
X
X
Soin(1968);
Altukhov(1979)
"Maynea" californica
4.0-5.3*
—
—
—
—
—
Kliever(l976)
Melanostigma atlanticum
-3.0-3.8*
—
—
—
—
—
Markle and
Wenner(1979)
Lycodes pacificus
5.0*
—
—
—
—
—
Levings(1969)
Lycodes palearis
6.0-7.0
Multiple
—
—
—
—
Slipp and DeLacy
(1952)
Lycodes pallidus
—
—
—
-10-12
—
—
Altukhov(1979)
Derepodichthys alepidotus
-2.1-2.4*
—
—
—
—
—
Anderson and Hubbs
(1981)
Bolhrocara hoUandi
9.2
—
—
35-36
X
X
Okiyama (1982a)
Bolhrocara sp.
7.0
1 (1.6 mm)
—
—
X (embryos)
X (embryos)
Kendall et al. (1983)
Lycodapus mandihularis
1.7-1.9*
1
—
-15-17
—
—
Anderson (1981)
Auslrolycus laticmctus
7.5-8.4*
—
—
-17
—
—
Gosztonyi (1977)
Dadyanos insigms
5.0
—
—
—
—
—
Gosztonyi (1977)
Phucocoetes latilans
4.5
—
—
20
—
—
Gosztonyi (1977)
Iluocoetes effusus
5.0-5.5
—
2
—
—
-
Gosztonyi (1977)
' Maximum ovanan diameters.
generally cursory (except that of Kendall et al., 1 983). In general,
zoarcid eggs are large (about 4-9 mm, except in some diminutive
species), spherical and usually with a single oil globule that may
have coalesced from a few smaller globules. Spawned eggs are
orange-yellow or purple with a somewhat darker orange or yel-
low oil globule (Anderson, 1981; Kendall et al., 1983) and have
a narrow perivitelhne space. Benthic egg masses are held to-
gether by a sticky, gelatinous mass that is not especially thick-
ened. Incubation times are known for only three species. Eggs
"hatch" in the ovary of Zoarces viviparus afler two months and
embryos develop for another two months therein (Fig. 307).
Embryos develop a dense vitelline vascular network that aids
in yolk resorption, respiration and assimilation of nutrients from
the mother's ovarian fluid (Soin, 1968). Olsen and Merriman
( 1 946) found that eggs of southern populations of Macrozoarces
americanus had an average incubation time of 2.5 months,
whereas eggs of northern populations took about 3.5 months to
hatch. Gosztonyi (1977) observed the eggs of Iluocoeles effusus
(as /. elongatus) from the Patagonian intertidal to require two
months to hatch during the austral autumn.
Larvae
As with observations on eggs, zoarcid "larvae" are not well
known, if this stage is developed at all. Although a few early
stages have been collected during ichthyoplankton surveys (Rass,
1949; Mattson and Wing, 1978; Altukhov, 1979), they are no-
tably absent in collections of other surveys where adults are
abundant, such as the Bering Sea (Musienko, 1963; Waldron
andVinter, 1 978) and offOregon (Richardson and Pearcy, 1977).
This is probably due to their short planktonic time. Early life
history stages of only five zoarcid species have been illustrated
(Kendall et al., 1983) and all these are reproduced here (Figs.
307, 308).
Females of Lycodapus mandihularis, Gymnelus viridis and
"Maynea''' californica^ are known to spawn larger eggs at in-
creasingly larger adult sizes, thus zoarcid hatching sizes vary.
In large eelpouts, like Macrozoarces and Zoarces. young hatch
at about 30-40 mm, but diminutive species, like Melagoslignia
and Derepodichthys are probably only about 1 0 mm at hatching.
At hatching, the yolk sac is rapidly taken into the gut. White
(1939) reported "the complete external disappearance of the
yolk" to occur in about 20 seconds in Macrozoarces americanus
that were stimulated to hatch in a pan of cold sea water. I have
observed a similarly rapid internalization of the yolk in larvae
of the liparidid Carcproclus sp. (Anderson and Cailliet, 1974).
Perhaps rapid yolk uptake is typical of fishes with a protracted
developmental period.
Newly hatched zoarcids strongly resemble adults. The major
difl^erences are the larger eyes and more rounded snout in the
young (Fig. 308). At the free-swimming, yolk-ingestion stage,
all fin rays have formed. The stage and direction of fin formation
in embryos is unknown in Zoarcidae. Most of the cephalic lat-
eralis pores were formed in larvae of Bolhrocara hollandi (as
Allolepis hollandi) noted by Okiyama (1982a). Post-hatching
Macrozoarces that I examined for this study had not developed
all their lateralis pores, a case similar to that of Gymnelus spp.
(Anderson, 1982). These planktonic young Macrozoarces had
absorbed their yolk and measured 33.8-36.0 mm SL. The young
fish were generally well ossified, except central regions of the
neurocranium and suspensorium. Jaw and pharyngeal teeth were
developed and a few had eaten copepods. In the smallest spec-
imen, the pectoral actinosts, scapula and coracoid were a fused
mass of cartilage, but these were separated and ossified in just
slightly larger specimens. Vertebrae were square in shape (rect-
' This species properly belongs in an unnamed, monotypic genus
(Anderson. 1984).
ANDERSON: ZOARCIDAE
A
581
Fig. 307. Zoarces viviparus. (A) egg and newly hatched embryo; (B)
developing embryo from mother's ovary; and (C) newly emerged young;
all from Soin (1968).
angular in adults) and all neural arches were fused, as in adults.
In the caudal skeleton, all fin rays and plerygiophores were
present, as in adults, but the neural arches of the first ural and
first preural centra were poorly developed, with some sections
free of the urostyle. Typical of many zoarcids, the caudal of
Macrozoarces has two epural. four upper hypural and 3-4 lower
hypural fin rays.
There are no specialized larval pigment patterns. The larvae
of Gymnelus viridis and Bolhrocara hollandi appear to be mono-
tone, as are most adults (Rass, 1949; Okiyama, 1982a). The
larva of Macrozoarces illustrated by White (1939) and those
examined by me bore the typical criss-cross pigment pattern of
older stages.
Meristic characters ofA/arroroarcfi early juveniles examined
fit within the range reported for adults (Table I 50). However,
Soin (1968) and Kendall et al. (1983) showed that developing
embryos oi Zoarces viviparus and Bolhrocara sp. had myomere
counts well below that of adult populations. Although large
sample sizes of most zoarcid genera are lacking for satisfactory
statistical analysis of meristic characters, the important thing to
note is that myomere addition seems to be a slow process in
zoarcids and that the full adult complement may not be reached
until embryos are very close to hatching. Alternatively, zoarcid
embryos and larvae may have differentiated myomeres with the
adult counts, but their small size and tight packing, particularly
near the tail tip, may make it difficult to observe them with a
conventional light microscope.
Fig. 308. Early stages of Zoarcidae. (A, B) Bolhrocara sp., after
Kendall et al. (1983); (C) Bolhrocara hollandi. after Okiyama (1982a);
(D) Gymnelus viridis. after Rass (1949); and (E) Macrozoarces ameri-
canus. after White (1939).
Relationships
The relationships of the zoarcids to other living fishes has
been confused in the literature. Greenwood et al. (1966) and
Rosen and Patterson (1969) allied the zoarcids to the gadiform-
ophidiiform lineages. Two of the four characters they used to
suggest this relationship, the presence of a basisphenoid bone
and free second ural centrum, both illustrated by Yarberry ( 1 965),
were shown to be erroneous by Anderson and Hubbs (1981).
Anderson (1984) suggested zoarcid relationships are within Gos-
line's (1968) Blennioidei, especially his superfamily Zoarceo-
idae. Eight of Gosline's 1 1 zoarceoid families were recognized
by Anderson (1984), with Lycodapodidae and Derepodichthyi-
dae synonymized under Zoarcidae and Stichaeidae expanded to
include Cryptacanthodidae and Neozoarcinae (see Makushok,
1 96 1 ; Peden and Anderson, 1978; Anderson and Hubbs, 1981).
582
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
A definitive phylogenetic reconstruction of zoarceoid rela-
tionships is not presently possible without a more thorough
knowledge of the anatomy of other fishes that have been tra-
ditionally allied with them. Preliminary phylogenetic inferences
were made by Anderson (1984), who also discussed relation-
ships among zoarcid genera. It should be noted, however, that
a search for more characters is still in progress. Makushok (1958)
and Springer (1968) suggested zoarceoid, or "northern blen-
nioid" relationships were not close to the "tropical blennioids,"
a fact that my own research supports. However, for the con-
venience of the reader, information on the early life history
stages of zoarceoids, excluding Zoarcidae, is given by Matarese
et al. (this volume) under Blennioidea, following Nelson ( 1 976).
Since there is a dearth of knowledge on early stages and since
the youngest specimens known of any zoarcid so closely resem-
ble adults, no early life history characters have helped in elu-
cidating systematic relationships within Zoarcidae, or the zoar-
cids to their allies. All the zoarceoids are characterized by
precocious early stages (see Matarese et al., this volume), but
the utility of these forms in phylogeny remains untested. Within
Zoarcidae, it is interesting to note that the development of ce-
phalic lateralis pores in the primitive Gymnelus viridis. Melan-
ostigma pammelas. and Macrozoarces americanus takes place
over a much longer growth period (up to 50-60 mm) than in
the more derived Bothrocara (Okiyama, 1982a) or in youngest
stages I studied of Lycenchelys (32 mm), Lycodapus (20 mm),
or Lycodes (38 mm). The value of this information awaits more
complete data on early life history stages of all zoarcids.
Virginia Institute of Marine Science, College of William
AND Mary, Gloucester Point, Virginia 23062. Present
Address: Department of Ichthyology, California
Academy of Sciences, Golden Gate Park, San
Francisco, California 941 18.
Gobioidei: Development
D. RUPLE
GOBIOIDS are one of the most speciose groups of fishes,
comprised of approximately 2,000 species or ten percent
of the total number of teleosts in the world (Cohen, pers. comm.).
Various workers have recognized from two to seven major fam-
ilies of gobioids, based on adult characters. For present purposes
I will recognize seven families' Eleotridae, Gobiidae, Rhyacich-
thyidae, Kraemeriidae, Gobioididae, Trypauchenidae, and Mi-
crodesmidae after Nelson (1976).
Development
Larvae are known for less than 5% of gobioid species. Eggs
and larvae are best known from Japanese waters (e.g., Dotsu.
1954, 1957, 1958, 1979; Dotsu and Fujita, 1959; Dotsu and
Mito, 1955, 1963; Dotsu and Shiogaki, 1971; Kobayashi et al.,
1973; Shiogaki and Dotsu, 1971e, 1972c). in the northeastern
Atlantic and Mediterranean Sea area (e.g., Petersen, 1917, 1919;
Fage, 1918; Lebour, 1919; Sparta, 1934; plus summaries in
Padoa, 1956f; and Russell, 1976), and less so in American waters
(e.g., Hildebrand and Cable, 1938; Perlmutter, 1939; Pearson,
1 94 1 ; and Ruple, in prep.). Most of these descriptive works deal
with the gobiids, although larvae are known for representatives
of all families except Rhyacichthyidae and Kraemeriidae.
Larvae of gobioids are fairly distinctive from other teleosts,
but considerable variation does occur within the suborder. The
diversity of characters found in eggs and larvae will be discussed
in the following section. This information was compiled from
published literature and the examination of gobioid larvae.
' Hoese (this volume) includes Gobioididae and Trypauchenidae in
the Gobiidae subfamily Amblyopinae and recognizes Xenisthmidae as
a distinct family. Eleotridae is changed to Eleotrididae.
Eggs
Eggs are known for eleotrids, gobiids, gobioidids, and micro-
desmids (Table 151). Eggs of eleotrids and gobiids are generally
ellipsoid and adhesive, many of which have filamentous strands.
Eggs range in size from as small as 0.40 x 0.32 mm in Eleotris
avycep/jfl/a (Eleotridae; Dotsu and Fujita, 1959) and 0.45 x 0.20
mm in Evorthodus lyricus (Gobiidae; Foster and Fuiman, MS
in prep.) to 3.8 x 1.3 mm in Pcrcottus glehni (Gobxiiisie:; Kry-
zhanovsky et al., 1951) and 5.5 x 0.9 mm in Acanthogobius
Jlavimanus (Gobiidae; Dotsu and Mito, 1955). Taenioides ruh-
icundus (Gobioididae) eggs are demersal, adhesive and measure
approximately 1.3 x 0.70 mm (Dotsu, 1957) whi\e Gunnellich-
ihys (Microdesmidae) eggs are spherical (Smith, 1958a).
Known gobioid eggs usually contain numerous small oil drop-
lets within the yolk. Newly hatched larvae range from 1.7 mm
in Aslerropteryx semipunctatus (Eleotridae; Dotsu and Mito,
1963) to 7.0 mm in Chaenogobius castanea (Gobiidae; Dotsu,
1954).
Larvae
Gross morphology. — Body shape of gobioids is generally slightly
elongate and slender, with body depth usually nearly uniform
rather than sharply tapering (Figs. 309-311). Gobioidid and
microdesmid larvae are moderately elongate and slender (Fig.
311), while most eleotrids and trypauchenids are only slightly
elongate and slender (Fig. 311). Microdesmids have the most
elongate body shape of any known gobioid larvae. Body form
within the gobiids exhibits the greatest variety, ranging from
fairly short and stout (Gobiidae Larva 1 , Fig. 309) to moderately
elongate and slender Luciogobius elongatus (Fig. 309). These
characteristic body shapes are usually retained from the larval
through adult stages.
RUPLE: GOBIOIDEI
583
Fig. 309. Larval gobiids from top to bottom: Gobiidae Larva 1. 3.0 mm NL [AMS (Australian Museum Sydney): JML82/ 1-2-2]; Luciogohus
elongalus. 1 2.0 mm SL (redrawn from Shiogaki and Dotsu 1 972c); Gobiidae Larva 2. 6.0 mm SL (AMS:JML 1 6-10-7); and Microgobius thalassinus
8.4 mm SL [GCRL (Gulf Coast Research Uboratory): 02035].
584
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 310. Larvae of gobiids from top to bottom: E.xpedio parvulns. 12.0 mm SL (redrawn from Shiogaki and Dotsu I971e); Astrabe lactisella,
11.1 mm SL (from Dotsu and Shiogaki 1971); Gobionellus beleosoma. 8.6 mm SL. (GCRL:02038).
The gut is generally straight and extends to about midbody
or just beyond (~50% to 65% SL) in most gobioids (Figs. 309-
311), although in many species the gut is slightly looped just
anterior to the vent as in Microgohius thalassinus (Fig. 309). In
the trypauchenid, Trypauchen microlepis (Fig. 311). the gut is
considerably shorter (~39% SL) than in other gobioids.
A prominent feature of gobioid larvae is a large gas bladder,
usually situated slightly anterior of midbody (Fig. 309). The gas
bladder is located just posterior to the pectoral fin in Trypauchen
microlepis (Fig. 311) and is smaller and less pronounced than
in most other gobioids. In small larval microdesmids ( <4.0 mm)
the gas bladder is located at about mid-gut, while in larger larvae
it is found about midbody, near the posterior portion of the gut
(Fig. 311). The prominent gas bladder in larvae usually disap-
pears by the juvenile stage, but is retained in the adults of some
species such as Gobiosoma atronasum (Colin, 1975).
Eyes of known gobioid larvae are basically round or slightly
ovoid in shape. The elongate gobioids such as the microdesmids
and gobioididshave small eyes (<20% HL) while most eleotrids
and gobiids have somewhat larger eyes (>20% HL).
The head is of moderate length (~ 16% to 34% SL), generally
slightly rounded and gently sloping. The shape of the head changes
drastically in many species as they transform into juveniles. In
microdesmids such as Microdesmus longipinms and Gunnelli-
chthys sp. the lower jaw becomes hooked and protruding during
the later pelagic larval stages (Fig. 311).
The lengths of dorsal and anal fin bases vary considerably
and are useful in the separation of gobioid larvae at various
taxonomic levels. The lengths of the fin bases are related to the
number of elements and/or the spacing between the individual
elements, which varies considerably. Trypauchenids, micro-
desmids, and some gobioidids, all have long dorsal and anal fin
bases (Fig. 311). Some eleotrids (Eleolris pisonis and Erotelis
smaragdus) and various gobiids (Rhinogobius similus. Yono-
RUPLE: GOBIOIDEI
585
'^^^^$?;;j^:::^xs$^\\\\\\l\
Fig. 311. Larvae of gobioids from top to bottom: Trypauchen microlepis (Family Trypauchenidae) 8.0 mm SL (ASMS:CFIT 2-11-78);
Microdesmus longipinms (Family Microdesmidae) 19.2 mm SL (GCRL:02036); Gobioides broussoniieti (Family Gobiodidae), 1 5.0 mm SL (GCRL:
02037); and Dormilalor macutatus (Family Eleotridae) 8.1 mm SL (GCRL:02039).
gobius boreus. and Luciogobiiis elongatus; Fig. 309) have shoin
fin bases, with few closely spaced elements.
Gobioids transform from larvae to juveniles over a wide size
range. The gobiids Gobiosoma bosci and G. robusluin begin
transformation at ~7.0 mm, while some microdesmids main-
tain a pelagic larval existence until they reach lengths of ~25-
35 mm.
Meristics and fin development. —Sequence of development,
number of elements, and size at which various fins develop are
586
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 151. Chae^acters Useful for the Separation of Gobioid Larvae to the Family Level. Characters present at least during postflexion
stage.
Larvae
Last fin to
Eggs
form and develop
Body shape
Gul length
Eye shape
Continuous or
separated dorsals
Dorsal and anal
fin base length
full complement
General
of elements
Eleotridae
Most ellipsoid
Slightly elon-
Midbody to
Round
Separated
Short to
First dorsal
and adhesive
gate and
slender
slightly be-
yond
long
Gobiidae
Most ellipsoid
Short and
Most to about
Round to
Most are
Short to
Pelvic
and adhesive
stout to
slightly elon-
gate and
slender
midbody or
slightly be-
yond
slightly
ovoid
separated
long
Rhyacichthyidae'
Separated
Short
Kraemeriidae'
Continuous
Long
Gobioididae
Demersal and
adhesive
Moderately
elongate and
slender
Midbody to
slightly be-
yond
Round
Continuous
Long
Pelvic
Trypauchenidae
Slightly elon-
gate and
slender
1/3 body
length to
slightly be-
yond
Slightly
ovoid
Contmuous
Long
Pelvic
Microdesmidae
Spherical
Moderately
elongate and
slender
Midbody to
slightly be-
yond
Round
Continuous
Long
Pelvic
' Larvae unknown, character states projected from adult conditions.
useful in distinguishing gobioid larvae at all levels. The sequence
of fin development is similar in most gobioids. although it vaiies
somewhat in eleotrids and gobioidids. Numbers of fin rays and
spines vary greatly among the gobioids and are particularly use-
ful in distinguishing gobioids at the family and species levels
(Hoese-). Degree of fin development at different sizes is helpful
in separating certain species of larvae, particularly if complete
developmental series are available.
Median finfolds and pectoral fins are present at hatching or
develop in early larvae of all known gobioids. The pelvic fin is
the last fin to form, usually beginning during the flexion or
postflexion stages.
The caudal fin is the first fin to form diflTerentiated rays, be-
ginning during the flexion stage. Gobioids usually have 17 seg-
mented principal caudal rays and numerous secondary rays that
are usually all differentiated by the postflexion or transforming
stages.
The second dorsal and anal fins are next to develop in eleo-
trids, gobiids, and microdesmids. The anterior and middle ele-
ments are first to form and development generally proceeds
posteriorly. It is difficult to distinguish the dorsal spines from
rays in the continuous dorsal fin of microdesmids during the
larval stages, but they are usually shorter than the rays. The first
element of the anal fin and the first element of the second dorsal
fin initially develop as rays, but later transform into spines in
most eleotrids and gobiids. The presence of a continuous (Afi-
crodesmus longipinnis. Fig. 311) or separated (Gobiidae Larva
2, Fig. 309) dorsal fin is useful in family diagnosis of gobioids.
This character varies considerably from the continuous dorsal
■ Hoese (this volume) provides a meristic table for gobioid families.
in microdesmids, gobioidids, and trypauchenids to the widely
separated fins of the gobiid Periophthalmus cantonensis (Ko-
bayashi et al., 1973).
The first dorsal and pectoral fins are usually next to complete
development in microdesmids and gobiids, while the full com-
plement of first dorsal spines is last to form in eleotrids. While
complete developmental series of eleotrids are sparse, it appears
that the posteriormost first dorsal spines form after the full
complement of pectoral and pelvic fin elements are present (e.g.,
Dormitator maculatus, Eleotris ptsonis. and Erotelis smarag-
dus). Inthegobioidid, Taenoides cirratus (9 .3 mm), Dotsu(1958)
depicts the pectoral fin to be the last fin to obtain its full com-
plement of elements. First dorsal spines range from 2 or 3 in
some gobiids (e.g., Claringer cosnmrus) to 28 in some micro-
desmids (e.g., Microdesmus longipinnis). The gobiid Luciogo-
bius elongatus (Fig. 309) lacks a first dorsal fin entirely (Shiogaki
and Dotsu, 1972c). Pectoral fin rays range from 3 to 26.
In known gobiids, trypauchenids, and microdesmids the pel-
vic fin is last to form and complete development. Development
of the pelvic fins in gobioids varies greatly between families and
within certain families such as the gobiids. Some gobioids have
strongly united pelvics that form a cup-shaped disc (Fig. 309)
at a very early age, while adult Rhyacichthys aspro (Rhyacich-
thyidae) have widely separated pelvics. Eleotrids, trypauchen-
ids, microdesmids, kraemeriids, and gobioidids usually have
separated or weakly united pelvics. Pelvic fins in gobiids range
from strongly united, forming a disc to weakly connected at the
base to totally separated (usually in species with reduced pelvics;
Fig. 309). The size at which pelvics develop is an important
character in the separation of some gobiid genera and species.
One pelvic spine and 2-5 rays occur in gobioids. Expedio par-
vulus (Gobiidae) lacks pelvic fins entirely (Fig. 310).
RUPLE: GOBIOIDEI
587
Table 151. Extended.
Larvae
Pelvic fin
Gas bladder
General bodv
Prominent ventral
condition
pigment
pigment
pigment
Separated
Dorsal surface
Moderate
Present
United to sepa-
Dorsal surface
Sparse to
Usually
rated
or dorsal and
posterior
surface
heavy
present
Separated
Usually sepa-
rated
Usually weakly
Dorsal and
Sparse
Absent
united
posterior
surface or
lacking
Usually weakly
Lacking
Sparse
Absent
united
Separated
Dorsal surface Moderate Present
Various other meristic characters are useful in the separation
of gobioid fishes. Branchiostegal rays number from 5 to 6. Myo-
meres/vertebrae range from 25 in Eviola infulata (Eleotridae)
to 76 in some of the microdesmids.
Pigmentation.— V\%men\a\\on on the gas bladder and along the
ventral surface of the body are considered to be characteristic
of most gobioid larvae. Melanistic pigmentation in gobioid lar-
vae varies considerably, from the heavily pigmented gobiid .-15-
trabe lactisella (Fig. 310) to the sparsely pigmented gobioidids
(Fig. 311) and trypauchenids. Larval gobiids, eleotrids, and mi-
crodesmids generally have a moderate amount of pigmentation.
Pigmentation patterns are especially useful in separating larvae
at the generic and specific levels.
Trypauchenid larvae, Ctenotrypauchen microcephalus and
Trypauchen microlepis (Fig. 311) generally lack pigmentation
except for two spots of pigment along the anterior poilion of
the caudal fin, while the gobioidids Taenioides cirratus and Go-
bioides hroussonneti totally lack pigmentation except for a pig-
mented gas bladder in Gobioides broitssonneti (Fig. 311).
Pigmentation on the gas bladder is a prominent feature of
most known gobioids. The most common condition is pigment
on the dorsal surface, as in Microgohius thalassinus (Fig. 309)
and appears in most known gobioids. Dorsal and posterior gas
bladder pigment is known only in Gobionclhis species and Go-
bioides broussonneti (Figs. 310, 311). Trypauchen microlepis
(Fig. 311), Ctenotrypauchen microcephalus. and Taenioides cir-
ratus are the only known gobioid species which lack gas bladder
pigmentation entirely during their early development.
The most pronounced pigmentation occurring in many eleo-
trids, gobiids, and microdesmids is that found along the ventral
surface of the body, in the region of the gut and anal fin base.
Along the anal fin base, this pigmentation often occurs on in-
ternal as well as external surfaces.
Pigmentation is often found in eleotrids, gobiids, and micro-
desmids on the caudal peduncle, along the dorsal surface of the
body, on the otic capsule, on the tip of the lower jaw, along the
mid-lateral posterior portion of the body, and on various fins.
While pigmentation often appears very similar, the subtle vari-
ations are frequently useful in the separation of larval gobioids.
Contribution of laiA'ae to systematics
Gobioid larvae have not been previously examined in terms
of contributing to the understanding of systematic relationships,
but I believe they will be of great use in the future. A preliminary
phenetic overview of gobioids based on characters available in
larvae (representing less than 5% of the total number of gobioid
species), presents some interesting groupings. Known larvae from
three eleotrid genera; Erotelis. Eleotris. and Dormitator seem
to form a cohesive group. Shared characters include: gross body
and head shape, short dorsal and anal fin bases, separated pel-
vies, gut length (~55% to 57% SL), dorsal gas bladder and
ventral pigmentation, late development of the first dorsal fin,
and separation of the two dorsal fins. Microdesmid larvae from
the genera Microdesmus. Cerdale. and Gunnellichthys. likewise,
all appear quite similar to each other, based on the following:
gross body and head shape, connected dorsal fins, long dorsal
and anal fin base, high vertebrae number, reduced pelvic and
pectoral fins, gas bladder and dorsal and ventral body pigmen-
tation.
While the above mentioned eleotrid and microdesmid groups
appear fairly cohesive as well as distinct from other gobioids,
the family Gobiidae seem to be in some respects a catch-all
group. Currently, many diverse types of gobies are included
within the family Gobiidae (some 250 genera and well over
1,000 species). It is possible that larvae may present us with
additional characters that may help to better define the group.
The use of only adult characters has led many workers to debate
the rank of many taxa, for example Nelson's (1976) families
Trypauchenidae and Gobioididae have been relegated to
subfamilial status within the Gobiidae (Hoese, this volume) or
lower by other workers. Although the use of larval characters
alone will not define gobioid families they may allow a better
understanding of relationships. Known larvae of Trypauchen-
idae and Gobioididae exhibit characters that are distinctive or
unique to these taxa. Trypauchen microlepis has the shortest
and most acutely looped gut of any gobioid and is one of only
three species that lack gas bladder pigmentation (others are a
trypauchenid and a gobioidid). Pectorals are also more reduced
than in other gobioids. The long continuous dorsal fin and long
anal fin base are not shared among most gobioids. Gobioides
broussonneti also has a long continuous dorsal fin and long anal
fin base. It is one of only two known gobioid genera with dorsal
and posterior gas bladder pigment (the other being the gobiid
genus Gobionellus). and is one of the most sparsely pigmented
gobioids known.
More descriptive work needs to be completed on the taxo-
nomic level of both adults and larvae before the full value of
ontogentic characters in gobioid systematics can be adequately
assessed.
Gulf Coast Research Laboratory, East Beach Drive, Ocean
Springs, Mississippi 39564.
Gobioidei: Relationships
D. F. HoESE
APPROXIMATELY 500 genera and 2,000 species of gobioid
fishes have been named. Currently, about 270 genera are
recognized, and it is estimated that the group contains between
1,500 and 2,000 species. About 50 families, subfamilies, and
tribes have been named. Gobioid fishes are distributed through-
out much of the tropical, subtropical and temperate regions of
the world, occurring in a variety of habitats in fresh, brackish,
and coastal marine waters to depths of about 220 meters on the
continental shelf Of the six extant families recognized here,
three (Eleotrididae, Gobiidae, and Microdesmidae) are world-
wide, and three (Xenisthmidae, Rhyacichthyidae, and Krae-
mariidae) are restricted to the Indo-Pacific. Most species of
gobioid fishes are benthic, but some are pelagic, many are bur-
rowers, and many live in burrows constructed by other organ-
isms.
Much of the early history of the classification of gobioid fishes
has been summarized by Iljin (1930), Koumans (1931), and
Miller (1973). Early classifications, based on external features
were provided by Gunther (1861), Bleeker (1874), Jordan (1923),
and Berg ( 1 940). Sanzo (1911) published the first extensive study
of the lateralis-system pores and papillae, characters which have
come into wide usage in the last 20 years at the generic and
specific levels. Regan (1911c) presented the first classification
based largely on osteological characters. He established the fam-
ily Psammichthyidae (=Kraemariidae), and provisionally placed
it with the gobioids, a placement which was not accepted until
relatively recently. The study of Regan was largely confined to
the cranial osteology, pectoral girdle, and vertebral numbers.
Gosline (1955) examined the osteology of a few representatives
of the major groups of gobioid fishes, and gave evidence for the
placement of microdesmids and kraemariids among the go-
bioids. Takagi (1950, 1953) contributed to the classification
based on examination of scales and the glossohyal, and later
(1966) published an extensive paper on the distribution of the
group. Akihito (1963, 1967) studied the scapula in a number of
species, and later (1969) presented one of the most detailed
studies of the higher classification of gobioid fishes, dealing with
major osteological features of 7 1 genera and 85 species, but did
not present a classification. Miller ( 1 973) described the osteology
of Rhyacichthys, and presented a largely new classification of
the group. Birdsong (1975) presented information on the prim-
itive character states for several osteological characters and in-
dicated presumed trends for each character. He also criticized
the classification of Miller and recommended a return to the
traditional classification.
Dawson ( 1 974b) characterized the Microdesmidae and (1973)
summarized distributional information on Indo-Pacific species.
Rofen (1958) reviewed the Kraemariidae. Matsubara and Iwai
(1959) described the osteology of Kraemaria scxradiata. Obrhe-
lova (1961) described a new family of gobioid fishes (Pirsken-
iidae) from Oligo-miocene fossil material from Europe.
Few studies have been carried out on the relationships of the
suborder to other fishes. Most early workers considered the
group related to perciform or scorpaeniform fishes. McAllister
(1968) and Freihofer (1970) suggested a relationship with the
Paracanthopterygii. Other workers have accepted a perciform
derivation (Miller, 1973: Springer, 1983; Gosline, 1955), al-
though Gosline (1971) suggested that the group might eventually
be regarded as a distinct order based on the structure of the
suspensorium and the caudal skeleton. No sister group has been
postulated.
Gobioid fishes are characterized by the following features: no
parietals; a pelvic intercleithral cartilage: interhyal displaced
away from the dorsal end of the symplectic: a gap between
symplectic and preoperculum: no orbitosphenoid or basisphe-
noid: lacrimal typically present, extending over maxilla, but not
forming lower margin of orbit: only one other suborbital rarely
present; fourth basibranchial cartilaginous; penultimate verte-
bra with a short expanded neural spine and an elongate ex-
panded hemal fused to centra; caudal skeleton with one to three
epurals, a small free parhypural, an enlarged lower hypural plate
articulating with and sometimes fused with urostyle. an enlarged
upper hypural plate fused to urostyle, and a small free upper
hypural; procurrent caudal rays articulate with cartilaginous
plates; lateral line usually absent on body, canals often devel-
oped on head, suborbital canal and mandibular canal usually
absent; first spine or ray, when spine absent, associated with
proximal elements of two pterygiophores (median element of
first pterygiophore of second dorsal fin rarely present); last two
rays of second dorsal and anal fins closely spaced and articulating
with a single pterygiophore in each fin. Meristics are given in
Table 152.
The following groups are recognized:
Rhyacichthyidae. — The monotypic family Rhyacichthyidae is
the most primitive gobioid fish in the following features: bran-
chiostegals 6; mesopterygoid and dorsal postcleithrum present;
lateral line present on body; an anterior sclerotic; lacrimal and
one additional suborbital present; 3 epurals; interhyal adjacent
to dorsal end of symplectic; 3 posttemporals; infraorbital and
mandibular head canals present; scales with multiple rows of
ctenii. Its specializations are related to adaptations to fast flow-
ing rocky streams and include: thickened muscular pelvic fins,
small mouth, placed ventrally and anteriorly.
Eleotrididae— The eleotridids, largely confined to freshwater
and estuarine environments, are currently definable on the basis
of the following primitive features: branchiostegal rays 6; pelvic
fins widely separate, pelvic girdle with a short post-pelvic pro-
cess, extending well beyond last pelvic ray, pelvic rays in line
with pelvic spine; mesopterygoid and dorsal postcleithrum gen-
erally present; interorbital normally broad; caudal peduncle long,
generally longer than second dorsal base; palatine normally more
or less L-shaped, with a short ethmoid process, articulating
medially with lateral ethmoid; scapula normally completely os-
sified; anterior sclerotic, suborbital (other than lacrimal), post-
temporals, and median element of first pterygiophore of second
dorsal fin usually absent; first basibranchial cartilaginous, ba-
588
HOESE: GOBIOIDEI
589
Table 152. Selected Meristics for Gobioid Families and Subfamilies.
Group
Branchi-
oslegal
rays
Dl
D2
PI
P2
Epurals
Segmented
caudal rays
Vertebrae
Rhyacichthyidae 6
Eleotrididae 6
Xenisthmidae 6
Microdesmidae
Microdesminae 5
Oxymetopontinae 5
Gobiidae
Oxudercinae 5
Amblyopinae 5
Sicydiinae 5
Gobiinae 5
Kxaemariidae 5
VII 1,8-9 1,8-9 21-22 1,5 3 17 12+16 = 28
III-X 1,6-17 1,6-13 13-21 1,5 1-2 15-17 10-18+11-19 = 24-36
0-VI 0-1,9-32 1,9-25 17-21 0-1.1-5 1-2 15-17 10-18+16-28 = 26-46
XX-XXVIII 26-66 23-61 10-15 1.2-4 1 15-17 42-76
VI 1,9-37 1,9-36 15-26 1,4-5 1 17 10-11 + 15-16 = 26
V-VIII 0-1,10-30 0-1,10-30 10-21 1,5 2 17 10+16 = 26
VI-VIIl 16-50 0-1,14-50 13-21 1,5 1-2 17 10+16-26 = 26-36
VI 1,9-11 1,9-11 15-23 1,5 1 17 10+16 = 26
0-X 0-1,5-19 0-1,5-19 11-25 1,4-5 1-2 13,16-17 10-16+14-21=25-36
IV-V 13-19 1,11-15 3-10 1,5 1 11 10-14+16-17 = 26-31
sibranchials 2 and 3 present; pterosphenoid present, corono-
meckelian bone present; pterygiophores of two dorsal fins nor-
mally continuous, without an intemeural gap. One group
(Leplophilypnits and Gohiomorphus and relatives) are special-
ized in having an intemeural gap (an intemeural space without
a pterygiophore) between the two dorsal fins. Members of the
group also have often lost several eleotridid primitive features,
such as the mesopterygoid and dorsal postcleithrum. Some are
specialized in having an unossified scapula and a single epural.
Other eleotridids consistently have 2 epurals and a well ossified
scapula. The group includes about 40 genera and the following
named taxa: Butinae, Belobranchinae. Gobiomoridae. Hypse-
leotrini, Milyeringidae, Ophiocarinae, and Philypni. Whether
any of these are recognizable must await further study.
Xenisthmidae— This coral reef group, restricted to the Indo-
Pacific, is treated extensively by Springer (1983) and is distinc-
tive in the following specializations: lower lip with a free ventral
margin; ascending process of premaxilla absent or rudimentary;
rostral ossified and functionally replacing ascending process of
premaxilla; first basibranchial ossified; basibranchials 2-4 ab-
sent; no pterosphenoid or coronomeckelian bone; intemeural
gap present between two dorsal fins. The two genera studied
lack the dorsal postcleithrum and the mesopterygoid. The group
includes 4 genera.
Microdesmidae— The group possesses the following primitive
features: maxilla more or less L-shaped, with a very short inner
process articulating medially with lateral ethmoid; usually sep-
arate pelvic fins, without an interspinal membrane. The group
is uniquely specialized in having a very long posterior pelvic
process. Other specializations include the strongly compressed
head and body, with lateral eyes; 5 branchiostegal rays; one
epural; dorsal postcleithrum and mesopterygoid absent. Trends
in the group include reduction of pelvic rays, the tendency for
the scales to become nonimbricate, and the development of a
very long-based second dorsal fin. Two subfamilies are recog-
nized here, but further studies may show both to be distinct
families.
Microdesminae.— The specializations include: maxilla with a
long strut-like anterior projection; body very elongate, with a
single dorsal fin attached to or reaching near caudal fin; dentary
with a long ventral process at anterior tip. The worldwide group
includes 5 genera and the following named taxa: Cerdalidae,
Gunnellichthyidae, and Paragobioididae.
Ptereleotrinae. — The specializations include: mouth almost ver-
tical; articular process of premaxilla absent or fused with as-
cending process; a single pterygiophore precedes the first hemal
spine. The worldwide group includes 6 genera (2 undescribed)
and the following named taxa: Nemateleotrinae. Pogonoculinae,
Oxymetopontinae. In addition both subfamilies of microdes-
mids contain several specializations sometimes found in Go-
biidae, such as the interlocking of the anterior preopercular
process with the dorsal end of the symplectic and the expanded
dorsal flange of the sphenotic reaching to the supraoccipital.
Gobiidae. — In some genera primitive features are found, such
as the ventral postcleithrum. 2 epurals, and separate head canals
between the eyes. Specializations include: pelvic fins usually
connected to form a cup-shaped disc, often separate in coral
reef genera, but interspinal membrane usually present; pelvic
spine displaced forward and ventrally, not in line with rays;
mesopterygoid and dorsal postcleithrum absent; palatine nor-
mally T-shaped, but L-shaped in some specialized genera; eth-
moid process of palatine extends across front of lateral ethmoid,
articulating with proximal base of lateral ethmoid or more com-
monly with median ethmoid; maxilla generally without an an-
terior process; median ethmoid displaced ventrally; an inter-
neural gap present between two dorsal fins (except in Trypauchen
and relatives. There may be one or two dorsal fins, and most
genera have 17 segmented caudal rays, rarely 13 or 16. Several
subfamilies have been recognized. Four are recognized here, but
further studies may considerably expand the number.
Oxudercinae.— Tongue fused to floor of mouth; a single pte-
rygiophore precedes first hemal arch; teeth flattened; second
dorsal fin usually long based; eyes displaced forward and up-
ward; 2 epurals, lateral process of sphenotic large and not in
contact with eye. The group occurs in mud and mangrove areas
in all tropical areas, except the New World. The group contains
about 10 genera and the following named taxa: Apocrypteidae,
Boleophthalminae. Periophthalmidae.
Amblyopinae.— Tongue fused to floor of mouth; 2 or 3 pteryg-
iophores precede first hemal spine; a single dorsal fin reaching
590
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
to near or fused with caudal; eyes rudimentary and placed for-
ward in orbit; lateral process of symplectic large; 2 epurals. The
worldwide group occurs in estuaries or off river mouths, and
includes about 10 genera and the following named taxa: Go-
bioididae, Taenioididae, Trypauchenidae.
Sicydiinae.— Tongue fused to floor of mouth or free only at tip,
highly modified jaw suspension; thickened and highly branched
pelvic rays and fleshy pads at tips of pelvic spines. The world-
wide group occurs in freshwater and includes about 5 or more
genera and the following included taxa: Sicydiaphiinae (in part).
Gobiinae.— The worldwide group includes about 200 genera
and is not easily definable. The group includes the following
named higher taxa: Aphyinae, Austrolethopinae, Benthophiliae,
Brachygobii, Calleleotrinae, Chaeturichthyi, Croilinae, Crystal-
logobiinae, Doliichthyidae, Gobiodontinae. Gobiosomini,
Gymnogobiini, Latrunculini, Lebetinae, Leioterinae, Luciogo-
biinae, Platygobii, Rhinogobiinae, Triaenophorichthyini, Tri-
dentigeriinae, Valencienninae.
Kraemariidae(=Psammichthyidae.)— The family agrees in most
features with the Gobiidae, being specialized in having a large
amount of cartilage in the skeleton and 3 pectoral radials. The
group is restricted to the Indo-Pacific and includes 2 genera.
Relationships
Most workers have generally agreed that the Rhyacichthyidae
and Eleotrididae represent the most primitive gobioid fishes
characterized by 6 branchiostegal rays, a mesopterygoid, and
dorsal postcleithrum. In addition other primitive features, not
found in gobiids are sometimes present, such as an anterior
sclerotic, lower suborbital (other than lacrimal), and extrascap-
ulae. Most other features generally retain a primitive nature in
eleotridids, such as 2 epurals, ossified scapula, head canals, when
present, separate between eyes, and a ventral postcleithrum.
Gobiids, microdesmids, and kraemariids have 5 branchiostegal
rays and lack a mesopterygoid and dorsal postcleithrum (with
over two thirds of the genera examined). These differences in
organizational grades have lead some workers to suggest that
the advanced gobiid level of organization may be polyphyletic
(Springer, 1983).
The primary innovate character defining the gobiid fishes is
the development of a pelvic cup-shaped disc, formed by mem-
branes connecting the inner pelvic rays and two pelvic spines
(interspinal membrane or frenum); with the forward and ventral
rotation of the pelvic spines on the pelvic girdle. It has been
shown that reef gobiids often have secondarily separate pelvic
fins (Hoese, 1971), although most species retain a rudiment of
the interspinal membrane and the typical gobiid pelvic spine
orientation. Consequently, the question of whether gobiid fishes
are monophyletic depends in part on whether the disc has evolved
independently in various gobiid groups. Studies of other gobiid
specializations, although incomplete, have not indicated that
gobiids are polyphyletic. For example Regan (191 1 c) first noted
that the eleotridids have an L-shaped palatine and gobiids a
T-shaped palatine. In eleotridids the ethmoid process of the
palatine is short and articulates directly with the middle of the
lateral ethmoid, while in gobiids the ethmoid process is typically
long, extending across to the median ethmoid, which is displaced
ventrally. Similarly in gobiids there is an intemeural gap be-
tween the two dorsal fins (a space between two neural arches
without a pterygiophore). Primitively in eleotridids, the pteryg-
iophores of the two dorsal fins are continuous, without a gap.
From the relationship between the pterygiophores of the second
dorsal and the anal fins, it appears that the gap in gobiids forms
from a posterior shift of the second dorsal fin. The intemeural
gap also occurs in Rhyacichthys and Xenisthmus, and several
eleotridid genera from New Guinea, Australia, and New Zealand
(Gobiomorphus, Philypnodon, Grahamtchthys, and two new
genera) and the Central American genus Leptophilypnus. Struc-
tural comparisons indicate that Rhyacichthys probably obtained
the gap by loss of a dorsal spine or forward shift of the posterior
dorsal spines. It is currently unknown whether the Xenisthmidae
and the Gobiomorphus- Leptophilypnus group are convergent
with gobiids or represent sister groups. Both groups sometimes
lose primitive eleotridid features such as the mesopterygoid and
dorsal postcleithrum. In general body form the Gobiomorphus-
Leptophilypnus group most closely approach the gobiid body
form expected of an ancestral gobiid. Although currently placed
with the eleotridids further studies are underway to determine
the relationships of the genera in the group.
The microdesmids also represent a gobiid level of organiza-
tion, in lacking several primitive features, but their relationships
to other gobioid fishes are unclear. The group is characterized
primarily by the unique specialization of having an elongate
posterior pelvic process. The two subfamilies have other spe-
cializations in common and show similar trends, with the Pter-
eleotrinae representing the primitive sister group. The micro-
desmids retain a primitive palatine-ethmoid articulation, and
the posterior pelvic process probably represents an elongation
of the short posterior pelvic process of eleotridids. Microdes-
mids share with gobiids the loss of the anterior branchiostegal
ray. The strong compression of the head may have lead inde-
pendently to the loss of the anterior branchiostegal ray. Unfor-
tunately no immediate sister group is known, although on the
basis of the intemeural gap, the Xenisthmidae represent a pos-
sible group. Although the inner rays of the two pelvic fins are
sometimes connected in microdesmines, no species is known
with an interspinal membrane. The microdesmines have pre-
sumably secondarily lost the intemeural gap. A similar situation
occurs in the gobiid Trypauchen, where a single long-based dor-
sal fin is present.
The kraemariids appear closest to gobiids. Whether the group
will remain a family is uncertain, since the group shows some
similarity to the gobiid Parkraemaria.
Since no immediate sister group has been postulated for go-
bioid fishes, relationships of the more primitive groups are un-
clear. Miller (1973) and Springer (1983) have recognized only
two gobioid families, Rhyacichthyidae and Gobiidae. Springer
(1983) has suggested that the Rhyacichthyidae represents the
primitive sister group of all gobioid fishes. It is clear that Rhy-
acicithys is more primitive than any other known gobioid (al-
though arguably only marginally more primitive than some eleo-
tridid genera, such as Micropercops), and at the same time
specialized. However, eleotridids do not show obvious inno-
vative specializations in relation to Rhyacichthys, but show loss
of some primitive features. Until a proposed phylogeny of prim-
itive genera becomes available, the eleotridids can only be re-
garded as a primitive stock, which gave rise to one or more lines
leading to the families recognized here. While most eleotridid
genera may well have evolved before the xenisthmid-micro-
HOESE: GOBIOIDEI
591
desmid-gobiid line (or lines) evolved, some genera, such as the
Gobiomorphus-Leptophilypnus group, may have evolved from
a common ancestor of the line (or lines).
Irrespective of the number of families, or subfamilies of go-
bioid fishes recognized, there is no obvious evidence to combine
the 40 eleotridid genera with any particular gobioid group. It is
clear that the interrelationships of this large group will not be
fully clarified in the near future.
The Australian Museum, 6-8 College Street, Sydney 2000,
Australia.
Scombroidei: Development and Relationships
B. B. COLLETTE, T. POTTHOFF, W. J. RICHARDS, S. UeVANAGI,
J. L. RUSSO AND Y. NlSHIKAWA
THE Scombroidei is a suborder of the Perciformes containing
6 families, 44 genera, and nearly 100 species. All species
are marine although at least one (Scomheromonts sinensis) moves
fairly long distances into fresh water. Most species are pelagic,
some epipelagic and some bathypelagic.
The first modem definition of the scombroid fishes as the
suborder Scombroidei was by Regan (1909). He clearly sepa-
rated the scombroids from such percoid families as the Car-
angidae, Rachycentridae, Coryphaenidae, Bramidae, and Men-
idae. Within the Scombroidei, Regan recognized four divisions:
I. Trichiuriformes (Gempylidae and Trichiuridae); II. Scom-
briformes (Scombridae); III. Luvariformes (Luvaridae); and IV.
Xiphiiformes (Istiophoridae, Xiphiidae, and three families
known only as fossils). Regan's Scombroidei was defined by
three primary characters: premaxillae beak-like, gill membranes
free from the isthmus, and epiotics separated by the supraoc-
cipital. To include Luvarus in the Scombroidei, reversals must
be postulated in these three characters. The relationships of
Luvarus lie with the Acanthuroidei (Regan, 1 902; Leis and Rich-
ards, this volume; Tyler et al., MS) and will not be considered
here. Recent workers have usually recognized a suborder Scom-
broidei that is essentially the same as that of Regan (1909)
including the Luvaridae (e.g.. Greenwood et al., 1966) or have
placed the billfishes (Istiophoridae and Xiphiidae), along with
the Luvaridae, in a separate suborder, the Xiphioidei (Gosline,
1968; Potthoff et al., 1980), or have removed Xiphias from the
group and placed it in its own suborder, Xiphioidei (Potthoff
and Kelley, 1982).
Scombroidei
Perciform fishes with epiotics separated by the supraoccipital,
gill membranes free from the isthmus, premaxillae beak-like,
upper jaw nonprotrusile (except in Scombrolabrax), predorsal
bones lost (except for a small one in Ruvettus, Thyrsites, and
Tongaichlhys and three well-developed ones in Gasterochisma).
second epibranchial extending over the top of the third infra-
pharyngobranchial (except in Gasterochisma), vertebrae 24 or
more, inlerorbital commissure of the supraorbital canals widely
incomplete or absent (Regan, 1909; Gosline, 1968; G. D. John-
son, pers. comm.).
Six families are recognized: Scombrolabracidae (monotypic;
Potthoff et al., 1980); Gempylidae (15 genera, 16 species; Mat-
subara and Iwai, 1952, 1958; Russo, 1983); Trichiuridae (9
genera, about 20 species; Parin and Bekker, 1972); Xiphiidae
(monotypic); Istiophoridae (3 genera, about 1 1 species; Naka-
muraetal., 1968; Morrow and Harbo, 1969; Nakamura, 1974);
and Scombridae (15 genera, 49 species; Collette, 1979, 1983).
Fig. 3 1 2 is a Wagner Tree based on 40 characters considered
significant in assessing scombroid relationships (see Appendix)
generated by the computer program (WAGNER 78) written by
J. S. Farris (following Farris, 1970; Farris et al., 1970). The tree
is rooted at Scombrolabrax which is considered as the most
primitive scombroid and was used as the outgroup for com-
parison with the other scombroid fishes. Numbers show where
a character changes from a plesiomorphous {Scombrolabrax
condition) to a derived-apomorphous state. The gempylids were
grouped together on this cladogram because data were not avail-
able on all the characters. The unresolved areas have been re-
solved in a separate study by Russo (1983) and are discussed
in the section on the Gempylidae. The cladogram shows several
synapomorphies of the billfishes and the Scombridae: pharyn-
geal tooth plate stay (character 3; G. D. Johnson, pers. comm.),
pair of small lateral keels at the base of the caudal fin (character
12), caudal fin rays covering hypural plate (character 14), etc.
Billfishes show many character reversals and independent ac-
quisitions. Within the Scombridae, most groups seem well-de-
fined.
Scombrolabracidae
From its original description by Roule ( 1 922), Scombrolabrax
heterolepis has been considered as related to gempylid fishes
(Grey, 1960; Gosline, 1968; Potthoff et al., 1980). In most in-
stances wherein Scombrolabrax differs from the gempylids it
differs in the direction of the percoids: upper jaw protrusile,
some opercular bones spinous or serrate, pelvic girdle relatively
strong and attached to the cleithra, no fusion in the caudal
skeleton, one fewer vertebra (17 + 13 = 30) than in any other
scombroid (except the billfishes) and procurrent spur present
but reduced (Gosline, 1968; Johnson, 1975; Potthoff etal., 1980).
The stay on the pharyngobranchial of the fourth gill arch that
is present in the Scombridae, Xiphiidae, and Istiophoridae is
absent as in the Gempylidae and Trichiuridae (Potthoff et al.,
1980). Roule (1922) originally placed Scombrolabrax in a sep-
arate suborder. Bond and Uyeno (1981) also recognized a sub-
order Scombrolabracoidei because of the unique specialization
in adults of the 5th through 12th vertebrae which are expanded
to form thin-walled bullae with wide ventral openings which
accommodate delicate bubble-like evaginations of the gas blad-
592
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
-o
a
• — ^
o
C/D
<
Q
OS
D
X
o
5
SCOMBRIDAE Qo] (g)
01
29113811^
Fig. 312. Wagner tree of scombroid fishes based on 40 characters (Appendix). Numbers are inserted where a character changes from a
plesiomorphous (Scombrolabrax condition) to an apomorphous state. Numbers in darlc circles show no homoplasy, those in light circles show
reversals in character state, and those in squares show independent acquisitions.
der. The presence of this autapomorphy is not sufficient reason
to place Scombrolabrax in a monotypic suborder. Taxa should
be grouped based on shared specializations.
Development
Scombrolabrax heterolepis (Fig. 313). — Larval development
was described by PottholTet al. (1980). Early larvae from 3-4
mm NL resemble the scombrid Thunmis in pigmentation, but
Scombrolabrax can be distinguished from Thunnus in having
only 30 myomeres as opposed to 39 myomeres in Thunmis.
Larger larvae acquire characteristic melanophores on the lower
jaw ramus and on the caudal peduncle.
Scombrolabrax shares characters with the Gempylidae and
the most primitive scombrid tribe Scombrini (Scomber, Ras-
COLLETTE ET AL.: SCOMBROIDEI
593
Fig. 313. Lateral views of scombroid larvae, (upper) Scombrolahrax heterolepis. 5.0 mm SL. from Potthotfet al. ( 1 980); (lower) Lepidocybium
JIavobrunneum 5.0 mm SL, western Atlantic, ATLANTIS II, Cr. 59, Sta. RHB2083, Nov. 26, 1970, drawn by J. Javech.
trelliger) (Table 153). Many of these shared characters are ple-
siomorphous and so are not useful in constructing a classifica-
tion.
Gempylidae
Body oblong or elongate, compressed; maxilla exposed; strong
anterior canine teeth present; base of spinous dorsal fin longer
than soft dorsal; three anal spines except Rexea and Nealotus.
with two spines; pelvic fins 1,5 or reduced to only a spine; caudal
fin present; vertebrae 32-58 (Tables 1 54 and 1 55); anterior pre-
caudal vertebrae without parapophyses, with sessile ribs; pos-
terior precaudal vertebrae with ribs attached at the extremities
of closed haemal arches (Regan, 1909). The family currently
includes 16 species in 15 genera (Parin and Bekker, 1972; Nak-
amura and Fuji, 1983; Russo, 1983).
Russo (1983) divided the Gempylidae into six monophyletic
groups (Fig. 3 1 4) based on osteological characters. Three groups
are monotypic: Lepidocybium. Rmettus. and Thyrsites. The
Epinnula group consists of four genera above character 7: Epin-
nula, Neoepinnula, Tongaichthys, and Thyrsitops. The Nealotus
group is composed of three genera above character 2: Nealotus,
Promethichthys. and Rexea. The Gempylus group contains five
genera above character 3: Thyrsitoides. Nesiarchus. Gempylus.
Diplospinus, and Paradiplospinus. Diplospinus and Paradiplo-
spimts should probably be combined under Diplospinus.
Development
The family Gempylidae is characterized by the following lar-
val and adult characters when compared to the family Scom-
bridae: known gempylid larvae (except Thyrsitops with smooth
spines) have serrate dorsal, anal and pelvic fin spines, scombrid
larvae have smooth spines (Table 1 53). Gempylids initially de-
velop 3 epurals (ontogenetic fusion in Diplospimis), scombrids
develop 2 epurals. Gempylids develop 2 uroneurals (we were
unable to confirm this on all gempylid genera), scombrids de-
velop one uroneural. In gempylids the first dorsal pterygiophore
inserts in the second intemeural space; in scombrids it inserts
in the third space. Most gempylids, except Ruveltus and Neoe-
594
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
i-
■H5
.1^
^
^
J
^
^^
^
^
.^
«0
-5f
f %
=:
o
s?
■*•
-5f
.•^
.5j
N,
V
"*-
^
<'
^
=0
^^
5j
/^
.4)
^"^
*
-V
CL^ i
=0
.'^
"if
Fig. 314. Most parsimonious cladogram of relationships of genera in the Gempyhdae (from Russo. 1983; fig. 47).
pinnula. have more precaudal and fewer caudal vertebrae; in
most scombrids the reverse is tnae (Table 1 54). In the gempylids
3 centra support caudal rays; in scombrids (except Scombrini
and Grammatorcynus) 4 or 5 centra support the caudal rays.
Lepidocybium flavobrunneum (Fig. 313). — Lepidocybiuin lar-
vae, caught in the Indian Ocean, have been described by Ni-
shikawa ( 1 982). The description agrees with the Atlantic Ocean
and Gulf of Mexico larvae from the MCZ collection examined
by Potthoff, except for the vertebral count which Nishikawa
reported to be 16 + 15 = 31. The MCZ specimens had 17 +
15 = 32 vertebrae. Lepidocybium larvae and juveniles can be
distinguished from other scorn brid and gempylid species by
meristics (Tables 154, 155, and 156), pigmentation and shape.
First dorsal fin spine count is the lowest for all gempylids and
scombrids. The first dorsal fin is intensely pigmented in larvae
of Lepidocybium and the individual spines have serrations. The
height and pigmentation of the fin is similar in larvae of Thunmis
and Euthynnus, but the fin spines are smooth in these two
scombrid genera. Neoepinnula has a considerably higher first
dorsal fin, also with serrate fin spines and also intensely pig-
mented. The low total vertebral count of 32 in Lepidocybium
is similar in Scombrolabrax, Scomber. Rastrelliger and Thyr-
sitops and the count is the same in Ruvellus, Epinmila and
Neoepinnula. The small projection on top of the head of Lep-
idocybium larvae as shown in Fig. 313 is also present in larvae
of Scomberomorus and Sarda. Cranial rugosities (striations)
observed in Lepidocybium larvae seem to be unique to this
genus. A very stout and long serrate preopercular spine is present
in Lepidocybium larvae. The overall intense gut pigmentation
Fig. 315. Lateral views of gempylid larvae from lop to bottom: Neoepinnula orientalis. 5.5 mm NL, Gulf of Mexico, Flower Garden 81-12,
Sta. 379, Nov. 8, 1981, drawn by J. Javech; Epinnula magistralis. 6.3 mm NL, modified after Gorbunova (1982); Thyrsitops lepidopoides,
5.5 mm NL, drawn from a specimen from Sato's (1983) study by J. Javech.
COLLETTE ET AL.: SCOMBROIDEI
595
596
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 316. Lateral view of the gempylid larva Thyrsiles atitn 6.0 mm NL, modified after Haigh (1972a).
in Lepidocybium is probably unique among scombrid and gem-
pylid larvae. In all known scombiid and gempylid larvae the
gut is intensely pigmented only dorsad with little lateral and
ventral pigmentation.
Lepidocybium has more affinities to the Gempylidae than to
the Scomberomorini. With the Gempylidae it shares caudal
skeletal characters such as 2 uroneurals and 3 epurals: scom-
brids have 1 uroneural and 2 epurals. Larval Lepidocybium have
serrate pelvic and first dorsal and anal fin spines, which also are
present in known gempylid larvae, except Thyrsitops. The first
dorsal pterygiophore in Lepidocybium. as in all gempylids, in-
serts in the second intemeural space in Atlantic specimens, but
in Pacific Lepidocybium the first dorsal pterygiophore is found
in the third intemeural space; in all scombrids the first dorsal
pterygiophore inserts in the 3rd space.
Ruvettus pretiosus.—T\\t larvae of Ruvellus are not known. This
lack of knowledge is surprising, because Ruvettus is caught as
by-catch in the tuna longline fishery (Nakamura, 1977). The
smallest Ruvettus known to us is 209 mm SL and has the features
of adults.
Epinnula magistralis (Fig. 315).— The larvae of Epinnula are
not well known. Gorbunova (1982) reported the capture of 3
larvae from the Gulf of Mexico and one larva from the Straits
of Yucatan. In larval Epinnula, the first dorsal fin is not as high
and not as intensely pigmented as in Neoepinnula and the first
dorsal fin is inserted more anteriorly in Neoepinnula than in
Epinnula. In Epinnula the preopercular spine is shorter than in
Neoepinnula. We believe that the 17.8 mm specimen figured in
Belyanina (1982b) is a specimen of Epinnula not Neoepinnula
because of the more posterior first dorsal fin insertion.
Neoepinnula orientalis (Fig. 3\ 5). — Neoepinnula larvae have
been described by Nishikawa and Nakamura (1978) and one
7.3 mm specimen was figured by Gorbunova (1982). Belyanina's
(1982b) figure of a 17.8 mm Neoepinnula probably is an Epin-
nula as mentioned above. The larvae of Neoepinnula are very
distinctive. They have a very high and intensely pigmented first
dorsal fin which inserts anteriorly almost on top of the head.
This causes the anteriormost first dorsal pterygiophores to insert
slanted in a posterior direction; no other gempylid or scombrid
larva has such a first dorsal fin.
Tongaichthys robustus. —Tht larvae of this recently described
genus and species are unknown (Nakamura and Fuji, 1983).
Thyrsitops lepidopoides (Fig. 315).— The larvae of Thyrsitops
were recently described by Sato ( 1 983). These are the only known
gempylid larvae which lack serrations on the fin spines. We
believe that the count of XVI-XXII first dorsal fin spines for
Thyrsitops given in Parin and Bekker's (1972) Table 4 is a
misprint and should be XVI-XVII.
Thyrsites atun (Fig. 316). — Haigh (1972a) described the larvae
of Thyrsites captured in plankton tows. Pigmentation is dis-
tinctive with 2 to 3 dark pigment blotches on the ventral tail
margin unlike any other known gempylid, but similar to the
trichiurid Benthodesmus. Haigh (1972a) gave counts for Thyr-
sites: XVIII-XXI first dorsal fin spines and 34-35 vertebrae.
Grey (1953) gave XX first dorsal fin spines and 37 vertebrae
and Parin and Bekker (1972) gave XX-XXI first dorsal fin
spines.
De Jager (1955) fertilized the eggs of a ripe Thyrsites female
with sperm from a male in the laboratory. The eggs hatched
and the larvae were fed drops of human blood. After 9 days
they died, visibly undernourished. De Jager illustrated the de-
velopment of the eggs and very early stages of the larvae. The
illustrations are not helpful for identification of wild caught
gempylid larvae because of starvation and underdevelopment.
The larvae figured by Regan ( 1 9 1 6) as Thyrsites are probably
Promethichthys or Rexea because the first dorsal fin in the figure
of the largest specimen shows XVIII spines and no pelvic fin
rays. Regan stated in the text that total vertebrae were 35.
Fig. 317. Lateral views of gempylid larvae from top to bottom: Promethichthys promelheus. 8.5 mm SL, modified after Gorbunova (1982);
Rexea solandri. 21.7 mm SL, Indian Ocean, DANA, Cr. 3915II1, from a cleared and stained specimen drawn by J. Javech; and Nealotus tripes,
9.0 mm SL, modified after Strasburg (1964).
COLLETTE ET AL.: SCOMBROIDEl
597
598
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
COLLETTE ET AL.: SCOMBROIDEI
599
Table 153. Comparison of Characters among Larvae and Juve-
niles OF Scombrolabrax, of known Gempylidae and of the Scombrid
Tribe Scombrini.
Scombrolabrax
Gempylidae
Tnbe Scombrini
First dorsal fin de-
after
before
after
velops before or
after second
First dorsal fin pte-
3
2
3
rygiophore inserts
in intemeural
space number
First dorsal, anal
smooth
serrated, ex-
smooth
and pelvic fin
cept smooth
spines of larvae
in Thyrsi-
serrated or
tops
smooth
Number of epurals
3
3
2
Number of uro-
2
2
1
neurals
Procurrent spur of
present
absent or pres-
absent
caudal fin (John-
ent reduced
son, 1975)
Hypural fusion
absent
present
present
Dorsal and anal
not bifur-
bifurcated
bifurcated
stay of posterior-
cated
most pterygio-
phore bifurcated
posteriorly or not
Rexea solandri (Fig. 3\1).—Rexea larvae are poorly known.
Parin and Bekker (1972) reported a 10.5 mm larva but did not
describe it. Six cleared and stained Rexea (21.7-28.9 mm SL)
from the "Dana" collections were identified by PotthofT from
meristics, but pigment was lost due to the age of specimens
which had been collected in the 1920s. In general, Rexea larvae
look similar to Promethichthys larvae. The pelvic spine is short-
er and does not reach the anus in Rexea, whereas in Prometh-
ichthys the spine reaches past the anus. Also, Promethichthys
has 3 elements associated with the first anal pterygiophore: Rex-
ea has only 2. All our Rexea larvae had one long serrate spine
and a miniscule vestige of a ray in each pelvic fin. The larvae
of Rexea promethoides are not known. Adult R. promelhoides
have a fully developed pelvic fin with a count of 1,5 rays.
Promethichthys prometheus (Fig. 317). — The larvae and young
of Promethichthys are poorly known. Giinther (1889) described
and figured two larvae 5 mm and 10 mm as Thyrsites prome-
theus (=Promethichthys prometheus). These larvae are not Pro-
methichthys. The smaller one cannot be positively identified but
could be a serranid larva because of body shape and number of
myomeres. The larger specimen is definitively Diplospinus mul-
tistriatus. Roule and Angel (1930) described and figured two P.
prometheus larvae 6 mm and 10 mm. From their description
and figures it is impossible to confirm their identification. We
APHANOPODIN/e DipLymus
Fig. 319. Relationships of the Trichiuridae and the gempylid
subfamily Gempylinae (from Tucker, 1956: fig. 23).
do not think they are Promethichthys because the pelvic rays
are too long and well developed, and there are 2 1 first dorsal
fin spines on the larger specimen. The larger specimen may be
Nesiarchus. Gorbunova (1982) described and figured two P
prometheus larvae 3.9 mm and 8.5 mm from the northern Ca-
ribbean Sea. The smaller specimen has only dorsal gut pigment,
a high first dorsal fin and long pelvic spines reaching past the
anus. The larger larva has a very high, moderately pigmented
first dorsal fin and a very long pelvic spine extending to the
anterior portion of the anal fin and a distinct pigment patch
near the hypurals. In body shape Promethichthys larvae resem-
ble those oi Re.xea. The first dorsal fin spine count in Parin and
Bekker ( 1972) for Promethichthys is a printing error. Total ver-
tebral counts for this genus given by Grey (1953) and Matsubara
and Iwai (1958) are 33 and 34 respectively, but ours were 34
or 35. This difference between our counts and Grey's and Mat-
subara and Iwai's is probably one of methodology. We counted
the urostyle as the last vertebra.
Nealotus tripes (Fig. 3 1 7). —The larvae of Nealotus are not well
known. Liitken (1880) figured two advanced Nealotus larvae,
but we are not certain of his identification. Strasburg (1964)
described a size series of Nealotus from 9 mm-41 mm. The 9
mm specimen had fully formed fins and probably had attained
some juvenile pigmentation. Nealotus has a very long posterior
process in the first anal pterygiophore which is evident in larvae
as small as about 8 mm. This is an excellent character to separate
Nealotus from Nesiarchus. Nealotus and Nesiarchus can be dis-
tinguished by their pelvic fin ray count and by the number of
middle radials in the second dorsal and anal fins. In juvenile
Nealotus the middle of the three anal spines fuses lengthwise to
the posterior process of the first anal pterygiophore. Thus, in
adult Nealotus only 2 anal spines are visible. One of us (PotthofT)
obtained many vertebral counts from post-larval stages and
Fig. 318. Lateral views of gempylid larvae from top to bottom: Nesiarchus nasutus. 7.5 mm SL. Gulf Stream off Miami, Virginia Key, Cr.
Fl, June 30, 1982, drawn by J. Javech; Gempylus serpens. 5.6 mm NL, Gulf of Mexico, OREGON II, Cr. 117, Sta. 34521, May 22, 1981, drawn
by J. Javech; and Diplospinus multistriatus, 7.1 mm NL, no data, drawn by J. Javech.
600
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 154. Number of Total First Arch Gillrakers and Number of Vertebrae in the Genera of Scombroidei.
Family and genus
Number of
species
Total number of gillrakers
on first arch
Number of vertebrae
Scombrolabracidae
Scombrolabrax
Gempylidae
Lepidocybium
Ruveltus
Epinnula
Neoepinnula
Thyrsilops
Thyrsites
Rexea
2
Promethichlhys
Nealolus
Thyrsitoides
Nesiarchus
Tongaichlhys
Gempylus
Diptospinus
Paradiptospinus
Trichiuridae
Aphanopus
Benlhodesmus
8
Lepidopus
2
Emxyinetopon
Assurger
Eupleurogrammus
2
Tentortceps
Trichiurus
Lepturacanthus
Xiphiidae
Xiphias
1
Istiophoridae
Tctrapturus
5
Makaira
3
Islwphorus
1-2
Scombridae
Gasterochisma
1
Scomber
3
Rastrelliger
3
Grammatorcynus
2
Scomberomoms
18
Acanthocybium
1
Orcynopsis
1
Cybiosarda
1
Gymnosarda
1
Sarda
4
Allothunnus
1
Auxis
2
Euthynnus
3
Katsuwonus
1
Thunnus
7
4-5
1.(6-8)*
1,(2-4)***
4,(13, 14)***
1.(4-9)***
1.(10)***
1.(0-5)***
1.(0)***
1.(5,6)***
28
Present
10-16
15-18
10-22
Absent, small teeth found
on epibranchial in one
juvenile.
Rakers modified to nu-
merous toothpatches in
Istiophoridae.
0
29-51
33-66
14-24
1-27
0
12-17
12-15
11-14
8-27
72-80
36-47
29-47
51-63
19-43
13
15, 16
17
10, 11
12
12
11
13
12
12
20,21
23
13, 14
17, 18
13
18
12-14
17-19
16-32
21-36
30-32
31-33
17, 18
19-21
22-24
23-26
19
19
22-28
20-27
20
19
20
19
20
17-19
20
21
18, 19
20,21
30
16
15
31*
17
15
32**
16
16
32**
15
17
32**
16
16
32**
17
16
32**
20
15
35*
22
15
37**
19,20
14, 15
34**
18
16
34*
20,21
14
34, 35**
20-22
14-17
36-39**
19,20
14, 15
34**
20-23
13-15
33-37**
17
16
33
26
23
49*
24-26
23-25
48-50** Pacific
26-29
24-26
51-55** Atlantic
22-28
30-37
58-61**
33,34
34
67**
42-44
55-56
98,99
38-53
65-103
123-156
41
70-73
111-113
39,40
63-65
103-104**
43
86
129
32-41
125-151
157-192
39,40
123-128
162-168
32-35
124-130
159-162
26
24
24
24
43,44
31
31
31
41-56
62-64
37-39
47,48
38
43-55
39
39
37-39
41
39
• From Matsubara and Iwai (1958).
•• Vertebral counts mostly from PotihofTs unpublished data,
*•• Numbers m parentheses are large spmes emerging from toothpatches. The number one outside parentheses represents a large raker in the epi-ceratobranchial angle. Numerous toothpatches
are present. Dunng ontogeny gillrakers transform to toothpatches.
COLLETTE ET AL.: SCOMBROIDEI
601
Table 155. Numbers of Spines and Rays in All Fins of Scombroid Genera. Numbers in parentheses denote vestigial rays and were counted
on cleared and stained specimens only.
Number of fin spines
ind rays
Firsl
Second
Dorsal
Anal
Caudal
Dorsal
Ventral
Family and genus
dorsal
dorsal
tinlels
Anal
tinlels
Pccloral
Pelvic
secondary
Principal
secondary
Total
Scombrolabracidae
Scombrolahra.x
12
11, 15-16
0
III, 16-17
0
18-19
I, 5
8-9
9 + 8
9-10
34-36
Gempylidae
Lepidocybium
8-12
16-18
4-6
II. 10-14
4-5
15-17
I, 5
10
9 + 8
10
37
Rmettus
13-15
15-18
2-3
III. 12-16
2-3
14-15
1.5
10
9 + 8
9-10
36-37
Epinnula
15-16
I. 16-19
0
HI. 13-16
0
15
I, 5
10
9 + 8
10
37
Neoepinnula
16
I. 16-20
0
III. 17-20
0
13-16
1,5
9-10
9 + 8
9-10
35-37
ThyrsUops
16-18
I. 14-16
4-5
III, 14-15
5
14-16
I, 5
8-9
9 + 8
8-9
33-35
Thyrsiles
20-21
II, 10-11
5-7
II, 11-12
6-7
14-15
I, 5
Rexea
16-19
-II, 13-16
2-3
I, 12-16
2-3
12-15
0-1,(1)71,2-3
8-10
9 + 8
8-9
33-36
Promelhichlhys
17-19
I, 17-21
2
l-II. 15-17
2
14
1(1-2)
lO-U
9 + 8
10
37-38
Neatotus
19-21
I, 16-19
2
II. 15-19
2
12-14
1(1-2)
8
9 + 8
9
34
Thyrsitoides
18-19
11,8
8
III. 8
1
13-15
I, 5
9
9 + 8
9
35
Nesiarchus
19-22
I, 19-22
2-3
II
-III. 15-17
2-3
13
I, 5
8-9
9 + 8
8-9
33-35
Gempylus
26-32
-11, 10-12
5-7
II
+ 1, 10-12
5-7
12-15
1,3-4
8-10
9 + 8
9-10
34-37
Diplospinus
30-36
1,35-41
0
II, 29-35
0
14
I
4
9 + 8
5
26
Paradiphspinus
36-44
28-33
0
II, 25-30
0
13-14
I
Tongaichlhys
12
I, 14
6
III, 14
5
17-18
I, 5
9
9 + 8
11
37
Trichiuridae
Aphanopus
38-41
II, 53-57
0
II, 44-50
0
12
I. 1 (juv.)
Present
Benthodesmus
32-46
70-109
0
11,65-101
0
12-13
1, 1
5
9 + 8
5
27
Lepidopus
9
90-97
0
11,61-64
0
1(1)
Present
Evoxymetopon
10
77-86
0
11,56
0
11-12
1.(1-3)
7
9 + 8
6-7
30-31
Assurger
Total 122
0
11,80
0
12
1
Present
17
Eupleurogra minus
3
123-147
0
11. 114-121
0
1,2
Absent
Tentonceps
Total 12
0
0
0
Present
Absent
Trichiurus
3
120-140
0
II. 105-108
0
Absent
Absent
Lcpluracanlhus
4
105-134
0
11.72
0
Absent
Absent
Xiphiidae
Xiphias adults
38-45
4-5
0
(12-
-16) + (3-4)
0
17-19
0
8-10
9 + 8
9-11
34-38
Xiphias
juveniles*
Total 44
-49
0
16-19
0
16-19
0
8-10
9 + 8
9-11
34-38
Istiophoridae
Tetrapturus
38-55
5-7
0
(11-
-19) -1- (5-8)
0
16-22
1,2
Makaira adults
38-46
6-8
0
(13-
-18) + (5-7)
0
18-23
1,2
13
9 + 8
13
43
Makaira
juveniles*
Total 5 1
0
25
0
18-23
1,2
13
9 + 8
13
43
Isliophorus
adults
37-49
6-8
, 0
(8-
-16) + (5-8)
0
17-23
1,2
11-12
9 + 8
11-12
39-41
Isliophorus
juveniles
Total 53
-54
0
23-24
0
17-23
1.2
11-12
9 + 8
11-12
39-41
Scombridae
Gasierochisma
17-19
9-10
6-8
10-13
5-8
19-22
I, 5
Scomber
9-13
11-12
5
I, 11-12
5
19-21
1,5
10-11
9 + 8
10-12
37-39
Rastretliger
9-11
12
5
I, 12
5
19-20
1,5
10
9 + 8
10
37
Grammalorcymis
11-13
10-12
6-7
11-13
6-7
22-26
1,5
Scombcromnnis
12-22
15-25
6-11
15-29
5-12
19-26
I, 5
11-13
9 + 8
11-13
39-43
Acanlhocybium
23-27
12-16
7-10
12-14
7-10
22-26
1,5
Orcynopsis
12-14
12-15
7-9
14-16
6-8
22-23
1,5
Cybiosarda
16-18
17-19
8-10
15-17
6-7
23-24
1,5
Gymnosarda
13-15
12-14
6-7
12-13
6
25-27
I, 5
Sarda
16-23
13-18
7-9
12-17
5-8
23-26
1,5
15-16
9 + 8
16-17
48-50
Allothunmis
15-18
12-13
6-8
13-14
6-7
24-26
1,5
Auxis
10-12
10-12
7-9
11-14
7
23-25
I, 5
15
9 + 8
16
48
Euthynnus
13-17
11-13
8-9
11-15
7-8
25-29
I, 5
15-16
9 + 8
14-16
47-49
Kalsuwomis
14-16
14-16
7-8
14-16
6-8
26-28
I, 5
16-17
9 + 8
17-18
50-51
Thunnus
11-14
12-16
7-10
11-16
6-10
30-36
1,5
15-17
9 + 8
15-17
47-51
• Xiphias and Isliophondae postlanae and juvenile have a continuous dorsal and anal fin.
602
ONTOGENY AND SYSTEM ATICS OF FISHES- AHLSTROM SYMPOSIUM
Fig. 320. Lateral views of trichiurid larvae from top to bottom; BenOwdesmus sp.. 8.1 mm NL. Gulf of Mexico, OREGON 11. Cr. 1 13, Dec.
6, 1980, drawn by J. Javech; Tnchiitrus lepturus. 6.3 mm NL and 17.0 mm NL, Gulf of Mexico, OREGON H, cruise unknown, Dec. 12, 1979,
drawn by J. Javech; and Lepidopus caudatus. 9.0 mm NL and 12.0 mm SL, modified from Padoa (1956a).
COLLETTE ET AL.: SCOMBROIDEI 603
Table 156. Characters for Gempyud Larvae, Juveniles and Adults and Our Knowledge of Gempylid Larvae and their Occurrence.
Species
Elements
associated
with first
anal pteryg-
iophore
Middle
radials.
Dorsal
Ventral
Dorsal and anal slay,
one or two parts, posteriorly bifurcated
Predorsal
bones
Larvae known
or unknown
Occurrence
Lepidocybium flavobrunneum
3
7
5
one part bifurcated
0
known
worldwide
Ruvcllus prcliosus
3
2,3
2,3
1
two parts bifurcated
1
not known
worldwide
Epinnula magistralis
3
two parts
0
poorly known
worldwide
Neoepinnuta orienlalis
3
0
0
4
two parts bifurcated
0
known
worldwide
Thyrsilops lepidopoides
3
two parts bifurcated
1
known
offcast and west
4,5
coasts of South
America
Thyrsites atun
—
—
—
—
known
all southern oceans
20"'-50°S
Rexea spp.
2
2,3
2,3
one part bifurcated
0
not known
Indian and West
Pacific oceans
Promelhuhylhys prometheus
3
2
2
two parts bifurcated
0
poorly known
worldwide
Nealotus tripes
3*
2
2
two parts bifurcated
0
poorly known
worldwide
Thyrsiloides marleyi
-
8
7
two parts bifurcated
0
not known
Indian Ocean, West
Pacific
Nesiarchus nasutus
3
3
3
two parts bifurcated
0
known
worldwide
Tongaichthys robuslus
-
6
5
two parts bifurcated
1
not known
Tonga Ridge
Gempylus serpens
3
6,7
6, 1
one part bifurcated
0
known
worldwide
Diplospmus mutlislnatus
3
1
one part bifurcated
0
known
worldwide
Paradiplospinus gracilis
3
1
1
one part bifurcated
0
poorly known
temperate and
arctic waters of
Southern Hemi-
sphere oceans
• Nealotus has .^ fin spine-ray elements associated in larvae and juveniles. The middle element gradually fuses to the long posterior process of the first anal pterygiophore.
juveniles of Nealotus. The great variability in vertebral counts
suggests that more than one species exists in the genus Nealotus.
Thyrsiloides marleyi. — Thyrsiloides larvae are unknown. This
is not surprising since the adults are considered to be rare (Nak-
amura. 1980).
Nesiarchus nasutus (Fig. 318).— The larvae of Nesiarchus are
well known. Giinther (1887) described a 33 mm pre -juvenile
specimen. A size series of 26 specimens 5.1-23.5 mm NL or
SL from the Atlantic Ocean was described by Voss (1954). The
5.1 mm NL specimen shown in Voss (1954: fig. 3A) and iden-
tified as Nesiarchus is Gempylus serpens because of the mid-
lateral pigment stripe and the large number of myomeres as
indicated by the close spacing. Nesiarchus larvae are easily iden-
tified by a heavily pigmented gular membrane. Larvae larger
than 6 mm develop pigment in the hypural area and a distinct
pigment stripe from the tip of the snout to the eye. One of us
(PotthofT) obtained many vertebral counts from post-larval
specimens and juveniles of Nesiarchus. The great variability in
vertebral counts suggests that more than one species exists in
the genus Nesiarchus.
Gempylus serpens (Fig. 318).— The larvae of Gempylus are
known. Liitken (1880) figured four post-larvae and juvenile G.
serpens. We believe that these were correctly identified because
at least 6 finlets are present on all but the smallest specimens.
Voss (1954) described 2 series of Gempylus larvae. Her Gem-
pylus A is Diplospinus and Gempylus B is G. serpens. Eight
larvae from 4.4 to 1 1.6 mm were described and the 5.1 mm
specimen in fig. 3A is a G. serpens not a Nesiarchus. Gempylus
serpens larvae can be distinguished from other gempylid larvae
by having a distinct line of lateral body pigment and up to 4
rays in the pelvic fin. The preopercular spines of Gempylus are
smooth, but the first dorsal and pelvic fin spines are serrate.
Late larvae and juvenile Gempylus develop 6 or 7 dorsal and
604
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
angustirostns belone
/
Ibidus oudox several species'^/ mdica nigricans
Table 157.
Comparison of Characters among Juveniles of
.\'/PH/4S AND IsTIOPHORIDAE.
Fig. 32 1 . Phylogenetic relationships within the family Isliophoridae
(from Robins and de Sylva (I960: fig. 5), names in quotations not
employed by Robins and de Sylva).
anal finlets and consequently have 6 or 7 middle radials. One
of us (Potthoff) obtained vertebral counts from late larval and
juvenile specimens from the Atlantic and Indo-Pacific oceans.
Total counts are higher for the Atlantic and lower for the Indo-
Pacific with a definite separation which may indicate that there
are separate species.
Diplospinus multisthatus (Fig. 318). — The larvae of Diplospi-
ntts are very well known, but some earlier researchers described
them under other names. The 1 0 mm specimen figured as Thyr-
sites prometheus in Gunther (1889) is Diplospinus because of
the anteriorly protruding spines on the tip of the lower jaw and
because of the distinct flatness of the ventral gut, although the
first dorsal fin spine count is too low for Diplospinus. Voss ( 1954
and 1 957) described Gempylus type A larvae, which definitively
are Diplospinus. Strasburg (1964) and Yevseyenko and Sere-
bryakov (1974) correctly identified and described Diplospinus
larvae.
The larvae of Diplospinus (Fig. 318) superficially resemble
those o{ Gempylus. but the larvae of Gempylus lack the following
characters present in Diplospinus: two horizontal spines at the
lower jaw tip, serrate preopercular spine, absence of pelvic fin
rays, flatness of ventral gut due to posterior process of basip-
terygium, and pigmented gular membrane. Larval Diplospinus
lack the lateral body pigment stripe characteristic oi Gempylus.
Diplospinus juvenWes lack dorsal and anal finlets and supporting
middle radials, features present in Gempylus.
Paradiplospinus gracilis.— Thelarvae of Paradiplospinus are not
well known. One of us (Nishikawa) has an unpublished manu-
script on the larval description.
Trichiuridae
Body elongate, strongly compressed; maxilla sheathed by
preorbital; anterior canine teeth strong; spinous dorsal not long-
er than soft dorsal (very slightly longer in occasional specimens
of .Aphanopus): two anal spines immediately posterior to the
vent; pelvic fins reduced to 1,1 or absent; caudal fin greatly
Xtphias
Istiophondae
Dorsal and anal fin
development, ad-
dition
First dorsal fin pte-
rygiophore inserts
in intcmeural
space number
Dorsal and anal stay
posteriorly bifur-
cated or not
Middle radial pres-
ent or absent for
posteriormost dor-
sal and anal pte-
rygiophore
Number of post-
cleithra
Pelvic fin and basip-
terygium present
or absent
Caudal fin rays sup-
ported by how
many rentra includ-
ing urostyle
Number of autog-
enous haemal
spines in hypural
complex
One pair of ribs on
centra
from a center in an
anterior and pos-
terior direction
not bifurcated
mostly from anterior
in a posterior di-
rection
1
bifurcated
absent
present
absent
present, fin ray num-
ber reduced
1-4 and 13-14
1-12
reduced or absent; dorsal spines and intemeurals correspond to
vertebrae, dorsal soft rays correspond to or are slightly more
numerous (in .Aphanopus and Benihodesmus) than vertebrae
(Table 155); vertebrae numerous, 98-99 (.iphanopus) to 192
(Eupleurogrammus) (Table 1 54); nbs feeble, sessile (Regan, 1 909;
Tucker, 1956). The family contains 9 genera and at least 18
species (Parin and Mikhailin. 1981). Most genera have only one
or two species; Benihodesmus has at least 8 valid described
species (Parin and Bekker, 1972; Parin, 1976, 1978).
Tucker (1956) recognized three subfamilies within the Trich-
iuridae (Fig. 319); Aphanopinae (Aplianopus, Benihodesmus,
and Diplospinus); Lepidopinae {Lepidopus. .Assurger. Tentori-
ceps, Evoxymeiopon, and Eupleurogrammus). and Trichiurinae
{Trichiurus and Lepturacanthus). Diplospinus and Paradiplo-
spinus have been transferred from a primitive position in the
Trichiuridae to an advanced position in the Gempylidae by
Russo (1983).
Development
Information on larval trichiunds is scarce. Of 9 trichiurid
genera only 3 species in 3 different genera have been described.
The known trichiurid larvae are characterized by very long bod-
ies, more than 100 myomeres, pelvic fins reduced or absent,
serrate spines in the first dorsal and anal fins and in the pelvic
fin if present. The first dorsal fin is the first fin to develop. The
Fig. 322. Lateral views of istiophorid larvae from Ueyanagi (1963a) rrom top to bottom: Istiophonis platypterus. 5.1 mm NL; Telraplurus
audax. 5.0 mm NL: T. angiistirosths. 4.5 mm NL; and Makaira mazara. 4.4 mm NL.
606
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 323.
Javech.
Lateral view of Xiphias gladius larva 6.1 mm NL, Gulf of Mexico, OREGON II, Cr. 126, Sta. 36784, May 25, 1982, drawn by J.
gut in preflexion larvae is visibly short, but elongates during
flexion and post-flexion.
Benthodesmus (Fig. 320).— Gorbunova (1982) described B.
Evseenko (1982b) described a 20.3 mm SL specimen. Pigmen-
tation in these larvae is strikingly similar to Lepidopus caudalus
larvae described by Padoa (1956a). However, in B. elongatus
simonyi the first dorsal fin spine is not more elongate than the
elongatus simonyi from a size series of 5 larvae 3.5-1 8 mm and other spines as in Lepidopus and Trichiurus (Fig. 320). Bely-
D
m
E
o
n
C
3
h
CO
0)
o
o
E
^
X
t/t
O)
o
u
i/>
y
k.
D
^
o
a
y
o
E
E
E
0)
-Q
E
c
o
c
o
O
o
o
o
o
o
u
to
D
(J
<
o
o
o
15
>-
u
D
-a
o
CO
o
-o
o
c
E
>-
O
3
C
C
3
C9
3
c
c
X
<
C
o
3
o
3
C
C
3
SCOMBRINI SCOMBEROMORINI
SARDINI
THUNNINI
T
GASTEROCHISMATINAE SCOMBRINAE
L-H
Fig. 324. Subfamilies, tribes and genera of Scombridae with number of species in each genus indicated (modified from Collcttc and Russo
1979: fig. 1).
COLLETTE ET AL.: SCOMBROIDEI
607
Fig. 325. Cladogram of relationships within the Scomberomorini (from Collette and Russo. in press).
anina (1982b) repoiled on a series of 22 B. elongatus pacificus
8.0-44 mm and on a B. vityazi 3 1 mm (Fig. 320). Benthodesmus
vityazi lacks dorsal pigmentation and has only two pigment
blotches ventrad, and the pelvic spine is reduced.
Lepidopus (Fig. 320). — The eggs and larvae of L. caudal us were
described by Padoa (1956a). The larvae are strikingly similar
to Benthodesmus in pigmentation; in Lepidopus the first dorsal
fin spine is longer than the following spines. Regan (1916) figured
an 1 1 mm larva as L. caudatus. It is impossible to determine
from the drawing and from the brief description if, in fact, it is
a larva of L. caudatus. The figured specimen is alcohol shrunk,
body and trunk pigments are absent, and the first dorsal fin
spine is shorter than the following spines.
Trichtunis lepturus (Fig. 320). — Delsman (1927) described
Trichiurus eggs and early larval stages hatched from wild caught
eggs. He believed that his descriptions were based on a number
of Trichiurus species. Newly hatched and early Trichiurus larvae
have a dendritic blotch of pigment, usually in the ventral finfold.
This blotch disappears when the first dorsal fin spines begin to
form anteriorly. Cutlass-fish larvae were also described by Gor-
bunova (1982) from a series of 59 specimens 5.0-1 7.2 mm and
by Tsukahara (1961) from a series of laboratory-reared and wild-
caught specimens. Small larvae lack pigment on the ventral
trunk and tail. With growth, a single row of melanophores ap-
pears just anterior to the first dorsal fin and develops posteriorly.
Ventral and lateral tail pigment is conspicuously absent even in
larger larvae. Trichiurus belongs to the tail-less trichiurids and
has no flexion stage. The pelvic fin in Trichiurus is absent.
Istiophoridae
Hypural plate mostly covered by caudal fin rays; caudal fin
supported by 3 centra (urostyle and preural centra 2 and 3); long
rounded rostrum formed by united premaxillae; nasals not
forming part of the bill; predentary bone present; teeth present;
pectoral fins placed low on body; scales present on body through-
out life; pelvic fins consisting of one spine and two long rays;
vertebrae few, (11-12) + (12-13) = 24; neural and haemal spines
expanded into strong overlapping laminae; ribs sessile (Regan,
1909; Gregory and Conrad, 1937). Three genera: Tetrapturus.
the spearfishes (six species). Makaira. the marlins (three species),
and Istiophorus. the sailfish (one or two species).
A diagram of relationships within the Istiophoridae was pre-
sented by Robins and de Sylva (1960) and is included here as
Fig. 321. Two additional species of Tetrapturus have been val-
idated since then: 7'. pfluegeri Robins and de Sylva and T
georgei Lowe. The former is most closely related to T. angus-
608
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 326. Lateral views of scombrid larvae from top to bottom: Scomber japonicus. 5.0 mm NL, modified after Kramer (I960); and Gram-
matorcynus bilineatus, 4.1 mm SL, modified after Nishikawa (1979).
tirostris and T. helone. the latter to T. albidus (Robins, 1974).
Nakamura (1974) recognized two species of blue marlins, the
Atlantic A/, nigricans and the Indo-Pacific M. mazura instead
of only M. nigricans. Morrow and Harbo (1969) considered
Istiophorus monotypic; Nakamura (1974) recognized the At-
lantic sailfish /. albicans as specifically, or subspecifically, dis-
tinct from the Indo-Pacific /. platypterus.
Development
Eggs. — No information is available on the identification of is-
tiophorid eggs, except for a brief account of eggs identified as
Tetrapturus belone by Sparta (1953).
Larvae.— Three studies, all of which appeared in 1974, sum-
marized the identification status of istiophorids (Richards, 1974;
Ueyanagi, 1974a, b). These larvae are extremely difficult to
identify. Two types of larvae are generally recognized— those
with short bills and those with long bills. The short-billed group
is generally referable to Makaira, the long-billed group to Is-
tiophorus and Tetrapturus (Fig. 322). Specimens less than 7.0
mm in length are all very similar. Other useful characters include
melanophore distribution on thegularand branch iostegal mem-
branes, relation of the pterotic and preopercular spines with the
body axis, shape of the orbit and position of the eye.
Meristic factors such as fin ray counts and vertebral formula
are not particularly useful in distinguishing istiophorid species
from each other (Richards, 1974; Tables 154and 155). Vertebral
counts can be used to distinguish Istiophorus and Tetrapturus
(12 + 12 = 24) from Makaira (11 + 13 = 24) at sizes greater
than about 20 mm (Richards, 1974). Probably the most useful
character is head morphology (Ueyanagi, 1963a). The snout is
short in all istiophorid larvae under about 5 mm in body length,
but in larger specimens the snout lengthens greatly in Istiophorus
and Tetrapturus. At lengths greater than about 1 2 mm, the
elongate snouts of Istiophorus and Tetrapturus readily distin-
guish them from the shorter-snouted Makaira. Thus, in ver-
tebral numbers and relative snout length, Istiophorus and
Tetrapturus are more similar to each other than to Makaira,
confirming the first subdivision in the family shown in Fig. 321.
For Pacific species, larval and juvenile stages are known for
all species except juvenile black marlin, M. indica. Makaira
indica larvae have a characteristic pectoral fin which is erect
from the body in larvae and adults and presumably juveniles,
too. Makaira mazara lai^ae are characterized by a short snout,
large eyes, and forward placement of the anterior edge of the
orbit. The characteristic lateral line appears in juveniles at about
30 mm SL. Tetrapturus aiida.x larvae do not have forward pro-
jecting orbits and the center of the eye is located at the same
level as the tip of the snout. The pterotic spine is parallel to the
body axis, and the preopercular spine is inclined sharply down-
ward, forming a large angle with the body axis. Melanophores
occur above the midline of the gular membrane or on the mid
COLLETTE ET AL.: SCOMBROIDEI
609
Fig. 327. Lateral views of scombnd larvae from top to bottom; Scomheromorus cavalla. 5.0 mm NL, eastern Atlantic. ALBATROSS IV. Cr.
7206, Sta. 79, capture date unknown, drawn by J. Javech; Acanthocybium solanderi. 7.2 mm NL. Gulf of Mexico, OREGON U, Cr. 1 17, Sta.
34463, drawn by J. Javech; Sarda sarda. 6.4 mm SL. Atlantic. GERONIMO. Cr. 3. Sta. 133. capture date unknown, drawn by J. .lavech;
and Gyinnosarda unicolor, 5.1 mm NL, modified after Okiyama and Ueyanagi (1977).
610
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 328. Lateral views oi Allolhunnus fallai larvae (upper) 5.7 mm NL and (lower) 6.9 mm SL modified after Watanabe et al. (1966).
anterior portion of the branchiostegal membrane. Istiophorus
platypterus has an elongate snout, small eyes, and a relatively
small head depth. Melanophores appear characteristically on
the posterior peripheral area of the gular membrane. However,
there are some sailfish larvae which lack gular melanophores,
and these are thought to belong to a different population. Tel-
rapturus angiistirosths larvae of small size are similar to M.
mazara, but the anterior edge of the orbital rim does not project
forward and melanophores occur on the branchiostegal mem-
brane.
In the Atlantic, specific differences are not nearly as clear.
Makaira indica larvae have not been identified although adults
are known. Makaira nigricans larvae are indistinguishable from
Makaira mazara and are characterized by the short snout, an-
terior projection of the orbital rim, and a lack of gular mela-
nophores. Tetrapturus pjluegeri larvae are very similar to T.
angustirostris and characteristically possess melanophores on
the branchiostegal membrane. It is also a winter spawning species,
whereas the others are spring and summer spawners. Tetrap-
turus albidus larvae are very similar to T. audax in the profile
of the head and possessing melanophores on gular membranes.
Tetrapturus georgei larvae are unknown, and T. he/one have
been briefly described without mention of the presence or ab-
sence of branchiostegal melanophores. A great amount of time
has been spent attempting to separate Atlantic /. platypterus
and T. alhidus with no success (Richards, 1974). Both have
relatively long snouts and pigmented gular membranes. Vari-
ation in gular and branchiostegal pigmentation has been de-
scribed in Atlantic T. pfluegeri (Ueyanagi, 1974b).
The elongate upper jaw, a characteristic of istiophorid fishes,
is also found in the fossils Palaeorhynchus and Blochius which
are thought to be the ancestral forms (Fierstine, 1972); hence,
the elongate upper jaw may have phyletic meaning. When the
character of upper jaw length compared to body length is ex-
amined during the larval period, clear differences were observed
(e.g., longer upper jaw in Istiophorus and Tetrapturus, and short-
er in Makaira). Adult T. angustirostris possess an especially
short snout among the species in the genus; elongation of the
Fig. 329. Lateral views of scombrid larvae from top to bottom: all drawn by J. Javech, Gulf of Mexico, OREGON II. Cr. 1 17, Auxis^p-. 5.0
mm NL, Sta. 34463; Euthynnus allelleralus, 6.2 mm SL, Sta. 34463; Kalsuwonus pelanus. 5.9 mm SL, Sta. 34448; and Thuimus thynnus. 6.0
mm SL, Sta. 34497.
COLLETTE ET AL.: SCOMBROIDEI
611
612
ONTOGENY AND SYSTEM ATICS OF FISHES- AHLSTROM SYMPOSIUM
Class intervals of numbers of coincident characters
state between pairs of genera
•:,(omber. !; , i,,-lh^:. ,
„ ,„„.,
Kuthynrtus; Katsuuoni. -.
'.
Sa rda
Acinlhocybiu"!
Group B
Fig. 330. Dendogram depicting larval relationships among 12genera
of Scombrinae (from Okiyama and Ueyanagi, 1978: fig. 2).
snout is greatest in the larval period. In the case of T. audax,
however, the snout/body length value is not so high among
Tetraplurus, and it is thought to be intermediate between Ma-
kaira and Tetraptums. The pattern of morphological change in
snout length with ontogeny in various genera corresponds to the
classification of adult fishes and is thought to reflect phylogeny
among the genera of istiophorid fishes (Ueyanagi, 1963b).
Table 1 58. Present Status of the Larval Fish Taxonomy of the
Family Scombridae (Modified after Okiyama and Ueyanagi, 1978:
TABLE 1).
Table 159. Presumed Phylogenetically Important Larval
(Chiefly Advanced Postlarval or Early Juvenile Stages) Char-
acters as Coded State for Comparison of 12 Genera of the
Subfamily Scombrinae. (After Okiyama and Ueyanagi, 1978: table 2).
Char-
Coded stale
index
Character
1
2
3
1
Supraoccipital
absent
—
present
spine
2
Head
small; less
than Vi
ofSL
large; more
than '/, of
SL
3
Viscera and
compact
—
elongated.
vent
with wide
space
from anal
fin
rounded
with vent
just in front
of anal fin
4
Snout
pointed
elongated
5
PremaxiUary
teeth
minute
large
large; some
fang-like
6
Jaw
equal size
equal or un-
equal size
unequal with
distinct up-
per jaw pro-
jection
7
Preopercular
spine
absent
—
present
8
Spiny supraor-
bital crest
absent
present or
absent
present
9
Pterotic spine
absent
—
present
10
Cartilaginous
pad on lower
jaw
absent
present or
absent
present
11
Dorsal body
pigmentation
heavier
—
lighter
12
Post vent pig-
present, ex-
absent or a
absent
mentation
tensive
few dots
13
Myotome
counts
low; 30-3 1
middle; 38-
41
high; 40-65
Subfamily, tnbe.
and genus
Present state of the lai^al fish taxonomy
Genenc level
Species level
Scombrinae
Scombrini
Scomber well established
RastreUiger well established
Scomberomorini
Grammatorcynus well established
Scomberomorus well established
Acanthocybium
Sardini
Orcynopsis
Cybiosarda
Sarda
Gymnosarda
Thunnini
Allothunnus
AlLXlS
Euthynnus
Katsuwonus
Thunnus
well established
no information
no information
rather well estab-
lished
well established
well established
well established
well established
well established
well established
well established
well established
well established
incomplete or none
for many species
well established
no information
no information
no problem in identi-
fication but poor
information
well established
well established
incomplete
well established
well established
established for most,
but identification
very difficult to ac-
complish
Xiphiidae
Hypural plate mostly covered by caudal fin rays; caudal fin
supported by only two centra (urostyle and preural centrum 2);
long depressed rostrum formed only by united premaxillae; na-
sals not forming part of the bill; predentary bone absent; su-
pratemporal absent; one postcleithrum; anteriormost dorsal pte-
rygiophore inserts in second (rather than third) intemeural space;
no teeth in adult; pectoral fins placed low; scales lost in adult;
pelvic fins and pelvic girdle absent; vertebrae few, 15-1- 11 =
26; neural and haemal spines not expanded; ribs present on only
the first four centra and the last two precaudal vertebrae (Regan,
1909; Gregory and Conrad, 1937; Potthofl'and Kelley, 1982;
G. D. Johnson, pers. comm.). Monotypic, contains only Xiphias
gladius.
Development
Xiphias gladius (Fig. 323). — Eggs and larvae of Xiphias have
been described by a number of authors during the 1 9th and 20th
centuries as summarized by Richards (1974). The most recent
and complete description is by Arata (1954) and drawings of a
developmental series are by Taning (1955). Osteological devel-
opment was studied by Potthoffand Kelley (1982).
Early larvae of swordfish are distinguished by having overall
body pigmentation and lacking the strong pterotic and pre-
opercular spines so characteristic of the istiophorids (Richards,
1974). Late preflexion larvae to juveniles acquire prickly squa-
I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 33 35 37 39
I I I I I I I I I I I I I I I I I I I I 1 I I I 1 I I I I I I I I I I I I 1 I
3 7 mm NL
-^
\j u M y u u y
Mtii
4 Omm NL
-^^QS^
0 0 0 0 a
l/VUUUUl/UUUV"""
i\ i\ r\ (\ n r\
4-4 mm NL
■^e:^^
'(? Po
UULI-UULXjUUui
3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Fig. 331. Osteological development of Thunnus atlanlicus. family Scombridae.
614
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
2 4 6 8 10 12 14 16 18 20 22
T r
-| r
~i r
"1 n I I I I r
-| 1 1 1 r—
3 0 mm NL
24
— I
4-
4 — e-
40mm NL
V/^
iXXXXXLXXXXx^^
4 3 mm NL
0 0 ff I) ff II ^ ffO d
4-5 mm NL
s /) 0 a ^ s n a ii 0 o a
4 9 mm NL
-\ — r-^^ — « — ^
\~i
0 0 € ff 0 oj 0 (I II 0 tP /? ^/j
L11UUU±SXS±±jU i I / iJTl
4 9 mm NL
■VTT^V"^^""^-^^^^
1^
6 6 mm NL
oOQooocdpOjo a (] ff ff a/j^ouoo 0 0 u g s a o « «„„„„.
^ ,j /, ,. ■ .
J^ OL. a^^^ i^
tXtWTTX^
4 6 8 10 12 14 16 18 20 22
Fig. 332. Osteological de\elopinent of Isliophonds platypterus, family Istiophoridae.
COLLETTE ET AL.: SCOMBROIDEl
615
Table 160. Developmental Features for the Scombroid Families and Morone. a Primitive Perctform Fish.
Scorn bro-
labracidae
Gempylidae
{Scombro-
(Gempvius.
Gempylidae
Thunnini
Scorn bnni
Isnophondae
Xiphiidae
Percichihyidae
lahrax)
Sestanhiis)
{Diplosptnus)
( Thunnus)
(Scomber)
(Isuophonts)
{Xtphia^)
i Morone)*
Developing neural
Antenor col-
Entire col-
Entire col-
Anterior col-
Not known.
Entire col-
Entire col-
Anterior
and haemal
umn: pos-
umn: pos-
umn: pos-
umn: pos-
umn: pos-
umn: pos-
column:
spines and arch-
teriorly.
tenorly.
teriorly.
teriorly.
teriorly.
teriorly.
posterior-
es and hypural
Center
Hypural
Hypural
Center
Hypural
Hypural
ly. Center
complex parts
column:
complex:
complex:
column:
complex:
complex:
column:
are added in a
anteriorly
anteriorly
antenoriy
anterioriy
anteriorly
anterioriy
anterioriy
direction.
and poste-
and poste-
and poste-
and poste-
and poste-
and poste-
and poste-
riorly. Hy-
riorly.
riorly.
rioriy. Hy-
riorly.
riorly.
norly.
pural
pural
Hypural
complex:
complex:
complex:
anteriorly
anteriorly
antenorly
and poste-
and poste-
and poste-
riorly.
riody.
rioriy.
Developing pter-
First dorsal:
First dorsal:
First dorsal:
First dorsal;
First dorsal:
Entire dor-
Entire dor-
First dorsal:
ygiophores and
antenorly
posterior-
posterior-
posterior-
very few
sal: very
sal: ante-
anteriorly
fin spines and
and poste-
ly. Second
ly. Second
ly. Second
antenorly.
few ante-
riorly and
and poste-
rays are added
riorly.
dorsal: an-
dorsal:
dorsal: an-
most pos-
rioriy.
posterior-
riorly.
in a direction.
Second
tenorly
some an-
tenorly
tenoHy.
most pos-
ly. Anal:
Second
dorsal: an-
and poste-
teriorly.
and poste-
Second
tenorly.
very few
dorsal:
teriorly
norly.
most pos-
riorly.
dorsal: an-
Anal: very
antenorly.
anteriorly
and poste-
Anal: an-
teriorly.
Anal:
terioriy
few ante-
most pos-
and poste-
riorly.
teriorly
Anal: few
some an-
and poste-
norly.
terioriy.
riorly.
Anal: an-
and poste-
anteriorly.
teriorly.
riorly.
most pos-
Anal: an-
teriorly
norly.
most pos-
most pos-
Anal: an-
teriorly.
teriorly
and poste-
teriorly.
teriorly.
tenorly
and poste-
riorly.
and poste-
riorly.
rioriy.
Sequence of fin
1. Second
1 . First dor-
1. First dor-
1. First dor-
1. Second
1. First dor-
1. Second
1. Second
and associated
dorsal
sal.
sal.
sal.
dorsal
sal.
dorsal
dorsal
pterygiophore
and anal
2. Second
2. Second
2. Second
and anal
2. Second
and anal
and anal
development.
concur-
dorsal
dorsal
dorsal
concur-
dorsal
concur-
concur-
rently.
and anal
and anal
and anal
rently.
and anal
rently.
rently.
2. First dor-
concur-
concur-
concur-
2. First dor-
concur-
2. First dor-
2. First dor-
sal. First
rently.
rently.
rently.
sal.
rently.
sal. First
sal. Sepa-
dorsal
First dor-
First dor-
First dor-
First dor-
dorsal
ration or
separated
sal sepa-
sal only
sal sepa-
sal nol
and anal
continui-
from sec-
rated
briefly
rated
separated
nol sepa-
ty of first
ond dor-
from sec-
separated
from sec-
from sec-
rated
and sec-
sal during
ond dor-
from sec-
ond dor-
ond dor-
during
ond dor-
part of
sal during
ond dor-
sal during
sal dur-
develop-
sals not
develop-
part of
sal dunng
part of
ing
ment.
given.
ment.
develop-
ment.
develop-
ment.
develop-
ment.
develop-
ment.
First anteriormost
Dorsal from
Dorsal from
Dorsal from
Dorsal from
Dorsal from
Dorsal and
Variable,
Dorsal and
dorsal and anal
one piece.
one piece.
one piece.
one piece.
one piece.
anal from
dorsal and
anal de-
pterygiophore
anal from
anal from
anal from
anal from
and anal
two
anal may
velop
develop from
two
two
two
two
from one
pieces.
develop
from two
one or two
pieces.
pieces.
pieces.
pieces.
piece.
from one
pieces.
pieces of carti-
or two
lage.
pieces.
Centra develop
Yes
Yes
Yes
Yes
Yes
No
No
No
from saddle-
shaped ossifica-
tions at bases of
neural and hae-
mal spines.
• Data from Fntzsche and Johnson (1980)
mation (Arata, 1954; Yabe et al., 1959; Potthoff and Kelley,
1982), and the scales are supposedly lost in adults (Palko et al.,
1981). One of us (Potthoff) has found that scales are retained
in adult swordfish, at least on some parts of the body. Larval
and juvenile swordfish differ from istiophoinds in a number of
characters (Table 157).
Adult Xiphias have two dorsal and two anal fins but larvae
and juveniles have single continuous dorsal and anal fins (Yabe
et al., 1959: fig. 9). During development, the fin rays in the
center of the fins stop growing and the rays become subcuta-
neous. The subcutaneous rays and their pterygiophores are pres-
ent in the adults (Potthoff and Kelley, 1982). In three scombrid
616 ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
Table 161. Osteological Features and Counts for the Scombroid Families and Morone. a Primitive Perciform
Fish.
Scombro- Gcmpylidae
labracidae {detnpylits.
{Scombrolahrax) Ncsiarchits)
Thunnini Scombnni Istiophondae Xiphiidae
iThimnus) {Scomber) {hliophonis) (Xiphtas)
Percichthyidae
i.\forone)*
Predorsal bones:
Present or absent absent
Number
0
usually ab-
sent#
Oor 1
First anteriormost dorsal pterygiophore;
Supports how 2 2
many spines
Inserts in inter- 3 2
neural space
number
First anteriormost anal pterygiophore:
Supports how 3 2 or 3t
many spines
Middle radials:
Present or absent present present
Dorsal and anal stay:
Present or ab-
sent, ossifies to part
one or two
parts, poste-
riorly
Bifurcated or non-bifur-
non-bifurcated cated
present: one present: usu-
ally 2
parts®
bifurcated
absent
0
2
2
2 or 3***
present or
absent
present: one
part
bifurcated
absent
0
2
3
absent
0
2
3
absent
0
3
1
present
present
present
absent
0
1 to 3,
mostly 2
2
1 to 3,
mostly 2
absent
present
3
3
3
present: one present: one present: one present: one
part part part part
bifurcated
bifurcated
bifurcated
non-bifur-
cated
present
present: one
part
non-bifurcat-
ed
Pelvic fin:
Spine, ray count
1,5
I, 5; I, 4; I,
2; I, 1:1
1, 3; I, 2; 1,
1; I
1,5
1,5
1,2
0
1,5
Preural centrum 3:
Neural spine
with or with-
out cartilage
tip
Haemal spine au-
togenous or
nonautogenous
with
autogenous
with
autogenous
with
autogenous
with
autogenous
with
autogenous
with
autogenous
without
non-autog-
enous
with
autogenous
Epurals:
Number
Anterior epural
fused with
neural arch of
PU,
3
No
3
No
1 to 3"
No
2
Yes
2
3
No
3
No
3
No
Uroneural:
1 or 2
2
2
1
1
1
1
1
2
Hypural 5:
Present or absent
Fused or separate
present
separate
present
separate
absent
fused early
present
separate
present
separate
absent
present
separate
present
separate
Ontogenetic hypural
fusion:
Present or absent
absent
present
present
present
present
present
present
absent
Procurrent spur (Joh
Present or absent
nson, 1975):
present
present, re-
duced or
absent
absent
absent
absent
absent
absent
present
Vertebrae inclusive of urostyle supporting caudal
Number 3 3
rays:
3
4
3
3
2
3
COLLETTE ET AL.: SCOMBROIDEI
617
Table 161. Continued.
Scombro-
labracidae
(Scomhrolahro-X)
Gempylidac
IGempylus.
Nestanhns)
Tnchiundae
Thunnini
( Thitnnus)
Scombnni
{Scomber)
Isliophondae
(Isriopfiorus)
Xiphiidae
(Xiphias}
Percichlhyidae
(Morone)*
Number of vertebrae:
Precau- 13+ 17 =
usually
usually few-
usually few-
13. 14 + 17,
12+12 =
15 + 11 =
12+ 13 = 25
da! + caudal = 30
more pre-
er precau-
er precau-
18 = 31
24
26
11 + 14 = 25
total
caudal,
dal, more
dal, more
11 + 13 =
16+10 =
fewer
caudal.
caudal.
24
26
caudal. 3 1
58-192
39-41
to 53
Stay on 4lh phar>ngobranchial (G.
D. Johnson, pers
comm.):
Present or absent absent
absent
absent
present
present
present
present
absent
" Data from Fnlzsche and Johnson ( 1 980) and G. D. Johnson, pers. comm.
# Rtivelltis. Thyrmops and Tongaichthys have one predorsal bone.
t Rexea has 2 spines, Nealolus ontogenelically has 3 spmcs but center spine fuses to basipterygium dunng development.
Si Leptdocybiuw. Rexea and Gempyltts have a one-pan stay.
•• Diplospinus ontogenelically has 3 epurals; poslenor 2 epurals are fused to one in adults, some Diplospinus develop only 2 epurals.
*•* Trichiunts has only 2 spines on the first anal pterygiophore.
genera. Scomber, Rastrelliger. and Auxis, there is a separation
between the first and second dorsal fins similar to that in adult
Xiphias, except that in these scombrids the two fins are separate
initially even though the first and second dorsal fin pterygio-
phores are continuous (Kramer. 1960),
Scombridae
Hypural plate mostly covered by caudal fin rays; caudal fin
rays supported by 3 centra (Scombrini and Grammatorcynus)
or 4 to 5 centra (all other tribes) (urostyle and preural centra
2 to 4); premaxillae beak-like, free from the nasals which are
separated by the ethmoid; no canine teeth; pectoral fins placed
high on the body, with 19-36 rays; pelvic fins 1.5; vertebrae 31-
64 (Tables 154 and 155); 5-12 finlets follow the second dorsal
and anal fins. The family contains 1 5 genera and 49 species
(Collette, 1979, 1983).
The family Scombridae can be divided into two subfamilies
(Fig. 324); the Gasterochismatinae, which contains only the
distinctive Gasterochisma nielampus, and the Scombrinae. There
are problems with the placement of Gasterochisma. To be in-
cluded in the Scombridae, it must have lost the extension of
the cartilaginous tip of the second epibranchial that extends over
the top of the third infrapharyngobranchial (Fig, 312, character
1; G, D, Johnson, pers, comm.), regain well developed predorsal
bones (character 2), and lose the pharyngeal tooth plate stay
characteristic of all other higher scombrids except Gasterochis-
ma and Grammatorcynus (G. D, Johnson, pers. comm.). How-
ever. Gasterochisma agrees with the billfishes and other scom-
brids in several caudal skeletal characters (Fig. 312. characters
8. 12, 14).
The Scombrinae is composed of two groups of tribes. The
primitive mackerels (Scombnni— Scomber and Rastrelliger) and
Spanish mackerels (Scomberomorini — 5cow/)<'roworj<.s, Acan-
thocyhium, and Grammatorcynus) have a distinct notch in the
hypural plate, lack any bony support for the fleshy keels on the
caudal peduncle, and do not have preural centra two and three
greatly shortened. The more advanced bonitos (Sardini) and
tunas (Thunnini) form a monophyletic group showing: loss of
the notch between the fused lower and fused upper hypural
bones (Fig. 312, character 33), bony support for the medial
caudal peduncle keel (character 1 6), anterior corselet of enlarged
scales (character 22) and have preural centra two and three
greatly shortened. The Scomberomorini. like the two more ad-
vanced tribes, have a median fleshy keel on the caudal peduncle
between the pair of small keels (character 1 1). However, there
is no bony support for this keel as there is in the bonitos and
tunas. Grammatorcynus shares this character state with the oth-
er two genera of Scomberomorini but has only three centra
supporting the caudal fin (reversal at character 8), as in the
Scombrini, rather than four or five as in the Scomberomorini,
Sardini and Thunnini. The Sardini (Orcynopsis, Cybtosarda,
Sarda, Gymnosarda, and Allothunnus: Collette and Chao. 1 975)
differ from the Thunnini (Auxis. Euthynnus, Katsuwonus. and
Thunnus) in lacking any trace of the subcutaneous vascular
system (Fig. 312. character 23) that permits the members of the
Thunnini to be warmer than the water around them. Instead of
being considered as a bonito. Allothunnus can better be inter-
preted as the most primitive member of the Thunnini. sharing
the presence of a prootic pit on the skull with the higher tunas
(character 26) but lacking their subcutaneous vascular system.
Allothunnus also has an autogenous second epibranchial carti-
lage as in the Thunnini (G, D, Johnson, pers. comm.) and shares
a common parasitic copepod, Elytrophora brachyptera. with six
of the seven species of Thunnus (Gibbs and Collette, 1967;
Cressey et al., 1983).
The Scomberomorini is the most speciose group within the
Scombroidei and so merits further attention. After comparing
the 18 species of Scomberomorus with each other and with
Acanthocybium and Grammatorcynus (Collette and Russo, in
press), characters that differentiated among species or genera
were listed. Grammatorcynus clearly is more primitive than
Scomberomorus and. therefore, it was used as the outgroup for
comparison with Scomberomorus. Character polarities were de-
termined by considering the character stale present in Gram-
matorcynus to represent the plesiomorphous condition. Of the
72 characters that differentiated at least one taxon from the
others, 14 were autapomorphies oi .Acanthocybium. These can-
not contribute to an understanding of relationships within
Scomberomorus and were omitted from the analysis. The re-
maining 58 characters were employed to generate a cladogram
(Fig. 325) using a computer program (WAGNER 78) written
by J. S. Farris (following Farris, 1970 and Farris et al., 1970).
618
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
The numbers at the nodes indicate characters that are discussed
by Collette and Russo (in press).
Scomberomorus differs from Acanthocybium at character 17
and from all other scombrids in possessing a spatulate vomer
that projects anteriorly well beyond the neurocranium. Scom-
beromorus differs from both Acanthocybium and Grammator-
cynus in 12 osteological characters. In three more characters,
Scomberomorus differs from both genera but is closer to Acan-
thocybium. Scomberomorus and Acanthocybium share 17 os-
teological synapomorphies at character 18 but differ from
Grammatorcynus. There are six species groups within Scom-
beromorus: sinensis, commerson, munroi. semifasciatus. gut-
tatus. and regalis (Fig. 325; Collette and Russo, in press).
The young stages of scombrids are difficult to identify to genus
and particularly to the species level (Richards and Potthoff,
1974). Young stages are, for the most part, easily identified to
family, but the eggs are unknown except for a few species. To
give some indication of the amount of work already directed to
these problems, a recent bibliography of young scombrids cov-
ering the years 1880-1970 listed 170 papers dealing with iden-
tification of eggs, larvae, and juveniles (Richards and Klawe,
1972). Where no specific references are indicated, information
is from papers listed by Richards and KJawe (1972) or Fritzsche
(1978).
Development
Scombrid eggs are very difficult to identify because they re-
semble the vast majority of perciform eggs characterized by 0.8-
1 .9 mm in diameter, smooth shell, usually a single oil globule
(several in Sarda), narrow perivitelline space, homogenous yolk
and a variety of distribution patterns of pigment cells including
melanophores and other pigments, usually yellow, white or green.
The latter three colors are lost upon preservation and are only
useful for identifying living eggs. Because of the great interest
in rearing scombrids from eggs, several papers have appeared
which describe living eggs, but not enough species have been
described to shed light on relationships. The following works
should be consulted: Harada et al. (1971); Mori et al. (1971);
Richards and KJawe (1972); Harada, Muruta and Miyashita
(1973); Harada, Muruta and Furutani (1973); Yasutake et al.
(1973); Harada et al. (1974); Ueyanagi et al. (1974); and Mayo
(1973).
Most larvae can be identified using a combination of char-
acters, principally number of myomeres, body shape, head spi-
nation, and distribution of melanophores. Larvae are unknown
for only three genera— Gasterochisma. Orcynopsis and Cybio-
sarda. The present state of knowledge of larval scombrids is
shown in Table 158. Morphological characters common to lar-
vae of this family are: (1) large head, large mouth opening and
large eye; (2) development of head spination; (3) posterior mi-
gration of anus (anus located in anterior region of body in early
larval stage; it migrates posteriorly toward anal fin during de-
velopment).
The following accounts follow the order of presentation in
Table 158.
Scomber and Rastrelliger (Fig. 326). — These two genera are
thought to be the most primitive and lack some of the larval
specialization seen in the other genera (Rastrelliger is not illus-
trated). The first dorsal fin forms after the second dorsal whereas
in other genera the first dorsal develops before the second dorsal.
The head is not large (less than 'A SL) in comparison with other
genera. The dorsal profile of the head is gently arched from
above the eye to the tip of the snout which is rounded. Head
spination is not developed. The typical pigmentation is the pres-
ence of melanophores on the mid- ventral side of the trunk and
tail in both genera. Myomeres number 3 1 . The species of Scom-
ber can be separated except it is difficult to distinguish S. ja-
ponicus from 5. australasicus. Head proportion and pre-anal
length may be useful as diagnostic characters.
Grammatorcynus (Fig. 326). — Grammatorcynus bilineatus lar-
vae resemble Scomber and Rastrelliger in dorsal profile of head
but have a pointed snout. Head spination is not developed but
preopercular spines are present. Typical pigmentation is the
presence of a lateral pigmented blotch above the anal fin and
the development of saddle-shaped pigment blotches on the body
and a pigment patch on the caudal fin base in larger larvae. Also
characteristic are two lateral lines which are discernible at 57
mm SL in juveniles. Myomeres number 31. The larvae were
recently re-described by Nishikawa (1979), but larvae of the
second species, G. bicarinatus. recently recognized by Collette
(1983) are unknown.
Scomberomorus (Fig. 327).— This speciose genus is character-
ized by having a supraoccipital protuberance (Euthynnus has a
slightly discemable protuberance). The head is large with an
elongate snout and large mouth. Preopercular spines are well
developed, and in at least one species, S. cavalla, are the longest
in the family. A spiny supraorbital crest is well developed. Me-
lanophores appear on the mid-dorsal and mid-ventral side of
the trunk and tail. Adequate descriptions have been published
for S. cavalla and S. maculatus and recently (Jenkins et al.,
1984) for 5. commerson. S. queenslandicus and 5. semifascia-
tus.
Acanthocybium (Fig. 327).— This single species has been well
described and is very easy to recognize (Wollam, 1969; Mat-
sumoto, 1968). It is characterized by a large number of myo-
meres (62-64), elongate gut, elongate snout, and melanophores
on the bases of the second dorsal and anal fins (on larvae >6
mm SL). It is the only species which does not exhibit posterior
migration of the anus.
Sarda (Fig. 327).— The snout is moderately elongate and the
head spination, consisting of supraorbital crests, preopercular
spines and pterotic spines, are well developed. Dentition on
both jaws is well-developed. Melanophores occurring on the
ventral midline appear to migrate dorsally along myosepta with
growth in a posterior to anterior direction. In postflexion larvae
the pelvic and first dorsal fin are heavily pigmented. Good,
thorough descriptions are lacking for all of the species.
Gymnosarda (Fig. 327).— The larvae of this monotypic genus
are unique in the remarkable development of the head, espe-
cially elongation of the snout, wide mouth with fang-like den-
tition, and spinous preopercles, supraorbital crests, and pterotic
spines. The extremes of the body proportions are: ca. 60% for
head in SL, ca. 60% for snout in head, and ca. 85% for upper
jaw in head. Melanophores are absent from the tail region and
the branchiostegal rays are heavily pigmented. The larvae were
described by Okiyama and Ueyanagi (1977).
COLLETTE ET AL.: SCOMBROIDEI
619
Allothunmis (Fig. 328).— This and the other genera of Thunnini
are very similar in appearance and are separated on the basis
of pigment patterns. All five genera have similar myomere counts,
preopercular spines present and spiny supraorbital crests absent.
Allothunrms fallai has 39 myomeres and unique melanophore
patterns are present on the mid-ventral surface of the lower jaw
along the base of the second dorsal fin.
Auxis (Fig. 329).— There appear to be two world-wide species
with 39 myomeres but there is some variation in pigment pat-
tern. The genus is characterized by having melanophores deeply
embedded behind the midbrain, cleithral symphysis, along the
ventral margin of the tail and melanophores absent from the
forebrain. The first dorsal fin is weakly developed and mela-
nophores occur along the lateral midline of the tail and on the
dorsal margin of the caudal peduncle in some specimens. The
profile of the head is blunt and the jaws are short giving the
larvae a characteristic ".'iMA/i-look" which is different from the
next three genera.
Euthynnus (Fig. 329).— Two species have 39 myomeres and a
third, E. lineatus, has 37. These larvae have slightly longer
snouts than other Thunnini and a slight supraoccipital protu-
berance. The unique pigment pattern is characterized by me-
lanophores occurring on the forebrain, midbrain, cleithral sym-
physis, and ventrally, laterally and dorsally on the tail. The first
dorsal fin is strongly developed and heavily pigmented.
Katsuwonus (Vig. 329).— The single species, K. pelamis. has 41
myomeres and a reduction in melanophores as they occur only
on the forebrain, midbrain, one to three distinct melanophores
on the ventral margin of the tail and rarely one or two on the
dorsal margin of the caudal peduncle.
Thunnus (Fig. 329).- All 7 species (Gibbs and Collette. 1967)
have 39 myomeres and show the greatest reduction in mela-
nophores in the family. Most species can be separated on the
basis of melanophores. Thunnus thynmis and T. inaccoyiihavt
melanophores on the ventral margin of the tail and the dorsal
margin of the trunk and tail. Thunnus ohesus and T. atlanticus
have melanophores only on the ventral margin of the tail. Thun-
nus alhacares and T. alalunga lack tail melanophores. Thunnus
tongol is unidentified. Geographic distribution, time of spawn-
ing and internal characters must be used to identify larvae of
this genus. We recommend that the following publications be
carefully consulted before attempting specific identifications:
Matsumoto et al. (1 972), Richards and Potthofr( 1 974), Potthoff
(1974, 1975) and Kohno et al. (1982).
Relationships
Okiyamaand Ueyanagi (1978) compared a classification based
on larval characters of 12 genera of Scombrinae with the clas-
sification of Collette and Chao (1975). They selected 13 pre-
sumed phylogenetically important larval characters (Okiyama
and Ueyanagi, 1978: table 2) and then coded the character states
(Table 159). Their dendrogram (Fig. 330) shows four groups.
Group A, Scomber and Rastrelliger. corresponds to the tribe
Scombrini (Fig. 324). Group B consists only of Gratnmator-
cynus. Group C equals the Thunnini (Fig. 324) plus Alloihunnus.
This interpretation is reasonable on cladistic grounds as dis-
cussed in the family section. Group D is a mixture of the Scom-
beromorini and Sardini. Okiyama and Ueyanagi admitted that
this group is a "heterogeneous assemblage."
The question of whether or not the billfishes should be con-
sidered scombroids has been addressed by Potthoff et al. (ms).
They studied osteological developmental features as shown in
Tables 160 and 161 and Figs. 331 and 332. Although their re-
search is still preliminary because of lack of adequate devel-
opmental series for many genera, they conclude that the Istio-
phoridae and Xiphiidae should not be placed within the
Scombroidei because of three developmental characters which
are not shared by any other scombroids. First, all scombroids,
except the Istiophoridae and Xiphiidae, have distinctive saddle-
shaped ossifications on the vertebrae before the centra are fully
formed. Second, development of the cartilaginous neural and
haemal spines also is similar in all scombroids, except istio-
phorids and xiphiids. Third, scombroids except istiophorids and
xiphiids share a primitive and an advanced development of the
first and second dorsal and anal fins and their supporting pter-
ygiophores. In the primitive development, which is shared by
Scombrolabra.x and Scombrini (and which is the basic devel-
opmental pattern of percoids), the second dorsal fin, anal fin
and pterygiophores develop first from a center anteriorly and
posteriorly and the first dorsal fin and pterygiophores develop
second, also from a center anteriorly and posteriorly. In the
advanced development, which is shared by the Gempylidae,
Trichiuridae and Thunnini, the first dorsal fin and pterygio-
phores develop first from the anteriormost element in a posterior
direction, and the second dorsal fin, anal fin and pterygiophores
develop second from a center anteriorly and posteriorly. In the
Istiophoridae, the first dorsal fin and pterygiophores develop
first from a center anteriorly and posteriorly. When the posterior
portion of the first dorsal fin development reaches above the
anterior portion of the anal fin, a few anal rays and pterygio-
phores develop anteriorly but most are added posteriorly. The
second dorsal fin develops only in a posterior direction consec-
utive to the first dorsal fin. In Xiphias the second dorsal and
anal fins and pterygiophores develop first from a center ante-
riorly and posteriorly. Development of the first dorsal fin and
pterygiophores then is continuous with the second dorsal fin
and in an anterior direction only.
In addition to their work, one can see the striking differences
between billfish larvae and other scombroids simply by review-
ing the illustrations of larvae in this report. However, these
synapomorphies of istiophorids and xiphiids are not shared with
any other group of fishes and so cannot be used as an argument
to relate the billfishes to any other taxa. Billfishes have another
unique synapomorphy: a specialized organ for heat production
located beneath the brain and adjacent to the eyes (Block, 1983).
The Scombridae, Istiophoridae and Xiphiidae have a stay on
the 4th phary ngobranchial that is absent in other perciforms (G.
D. Johnson, pers. comm.). Until further work is completed and
other characters thoroughly studied, the billfishes are retained
in the Scombroidei. The larval evidence presented indicates a
close relationship among the families Scombrolabracidae, Gem-
pylidae, Trichiuridae and Scombridae and much more distant,
if any, relationship to the Istiophoridae and Xiphiidae.
620
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Appendix
Characters and character states used for cladogram of scombroid fishes (Fig. 312). 1. Epibranchials. Tip of 2nd epibranchial short
(0, plesiomorphous); 2nd epibranchial extends over top of 3rd infraphar^ngobranchial to connect with 3rd epibranchial ( 1 , apomorphous).
2. Predorsal bones. Absent (1); present (0). 3. Pharyngeal tooth plate stay. Stay absent (0); stay present on 3rd pharyngeal tooth plate
where it contacts 4th pharyngeal tooth plate (1). 4. Dorsal fin developmental sequence. Second dorsal develops before first dorsal (1);
first dorsal develops before second dorsal (0). 5. Shape of premaxilla in larvae. Not beak-like (0), beak-like (1). 6. Cross-connections
of gill filaments. No cross-connections (0); cross-connections present (1). 7. Number of epurals. Three (0); two(l). 8. Number of vertebrae
supporting caudal fin. Two (0); 3-4 (I). 9. Infraorbital bones. Not expanded into large plates (0); expanded mto large plates (1). 10.
Subocular shelf Present (0); absent (1). 11. Mid-lateral keel on caudal peduncle. Absent (0); present (1). 12. Pair of small keels at base
of caudal fin. Absent (0); present (1). 13. Bill. Absent (0); present (1). 14. Extension of caudal fin rays over hypural plate. Not overlapping
or slightly overlapping plate (0); completely covering plate (1). 15. Anterior end of infraorbital bone. Not tubular (0); tubular (1). 16.
Bony keels on caudal peduncular vertebrae. Absent (0); present (1). 17. Bony caudal peduncle keels. Poorly or irregularly developed
(0); well-developed, forming a wide plate (1). 18. Inner row of premaxillary teeth. Additional row of teeth present on antero-medial
end of premaxilla (1); single row of premaxillary teeth (0). 19. Prolrusability of premaxilla. Upper jaw protrusible, premaxilla free from
maxilla (0); premaxilla anchored to maxilla (1). 20. Number of ossifications in last dorsal and anal pterygiophores. A single ossification
(0); two ossifications ( 1 ). 2 1 . Relationship of second dorsal fin pterygiophores to neural spines. Relationship 2: 1 (0); 1:1(1)22. Corselet.
Absent (0); present (1). 23. Subcutaneous vascular system. Absent (0); present (1). 24. Fronto-panetal fenestra. Absent (0); present (1).
25. Tooth shape. Conical (0); compressed (1). 26. Prootic pits (on ventral surface of skull). Absent (0); present (1). 27. Vertebral trellis
work. Absent (0); present (1). 28. Joint between first and second infraorbital bones. Simple contact (0). tightly bound (1). 29. Number
of vertebrae. Moderate numbers, 30-31 (1); few, 24-26 (0); many, 35-170 (2). 30. Number of uroneurals in adult. Two (0); one (1).
31. Fusion of uroneural to urostyle. No fusion (0); fused (1). 32. Fusion of upper hypural bones. Hypurals 3, 4, and 5 separate (0); 3
and 4 fused ( 1 ); 3, 4, and 5 fused (2). 33. Notch in hypural plate. Large (0); small ( 1 ); absent (2). 34. Fusion of upper and lower hypural
plates. Not fused (0); fused (1). 35. Fusion of lower hypural bones. Hypurals 1 and 2 separate (0); fused (1). 36. Fusion of parahypural
to hypural plate. Separate (0); fused (1). 37. Number of autogenous haemal spines. Two (0); one (I). 38. Tips of neural and haemal
spines of preural vertebra 4. Tips of both flattened (2); tip of one flattened (1); tips not flattened (0). 39. Number of pectoral fin rays.
17-19 (1, plesiomorphous); 10-17 (0, apomorphous); and along another transition series to 17-23 (2), 20-27 (3), 25-29 (4), and 30-
36 (5). 40. Tongue teeth. None fused to glossohyal (0); two patches fused to glossohyal (1).
(B.B,C, AND J.L.R.) National Marine Fisheries Service, Na-
tional Systematics Laboratory, National Museum of
Natural History, Washington, District of Columbia
20560; (T.P. and W.J.R.) National Marine Fisheries Ser-
vice, Southeast Fisheries Center, 75 Virginia Beach
Drive, Miami, Florida 33149; (S.U.) Faculty of Marine
Science and Technology, Tokai University, 3-20-1
Orido, Shimizu-Shi, 424, Japan; (Y.N.) Far Seas Fishery
Research Laboratory, Fisheries Agency, 1000 Orido,
Shimizu-Shi, Japan,
Stromateoidei: Development and Relationships
M. H. Horn
THE Stromateoidei is a suborder of perciform fishes com-
posed of six families, 16 genera and approximately 65
species. These fishes form a reasonably well-defined group that,
with one exception, is characterized by toothed saccular out-
growths in the alimentary tract immediately posterior to the last
gill arch. All species have small uniserial teeth in the jaws.
Stromateoids are marine fishes of temperate and tropical lat-
itudes and range from inshore coastal waters to the open ocean
and from pelagic (primarily) to demersal habitats. Of the 16
stromateoid genera, eight are exclusively oceanic, two are mixed
coastal and oceanic and six are exclusively coastal (Table 162).
Although several coastal species are locally abundant and com-
mercially important, oceanic stromateoids tend to be rare and
sporadic in occurrence. Juveniles commonly associate with an-
imate or inanimate floating objects in the surface layers of the
ocean.
Since the completion of Haedrich's (1967) comprehensive
review of the stromateoid fishes, several taxonomically-oriented
studies have been conducted. These works include descriptions
of a new monotypic family (Haedrich, 1 969) and four new species
(Haedrich, 1970; Horn, 1970b; Chirichigno, 1973; McAllister
and Randall, 1975), generic reviews or revisions (Horn, 1970b,
1973; Butler, 1979), regional reviews of certain centrolophid
taxa (Stehmann and Lenz, 1973; McDowall, 1982) and an ex-
tensive account of the early life history stages of pelagic stro-
mateoids (Ahlstrom et al., 1976). The present paper, which
includes a phylogenetic analysis of stromateoid genera, draws
heavily upon information in Haedrich ( 1967) and Ahlstrom et
al. (1976).
Development
Eggs
The eggs of approximately 14 species representing six genera
and four families of stromateoids have been described (Table
163). Stromateoid eggs typically are relatively small (0.70-1.80
mm in diameter), pelagic, separate and spherical. They have
unsculptured surfaces, unsegmented yolks and single oil glob-
ules. The few distinctive features of the eggs limit their value
as a source of taxonomic characters.
■^/-._
Fig. 333. Examples of stromateoid larvae and early juveniles. (A) Amarsipus carlshergi (Amarsipidae), 16.7 mm postflexion larva; (B)
Schedophilus hulloni (Centrolophidae), 25.0 mm early juvenile: (C) Icichthys lockingtom (Cenlrolophidae), 20.0 mm early juvenile; (D) Nomeus
gronovii (Nomeidae), 22.7 mm early juvenile; all from Ahlstrom et al. (1976).
622
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 162. Family Affiliation. Habitat and Ranges of Meristic Values for Stromateoid Genera.
Amarsipidae
Centrolophidae
Amarsipus
Hyperoglyphe
Cenfrolophus
Schedophihts
Icithlhys
Tuhhia
StTiolella
Habitat
Coastal
X
X
Oceanic
X
X
X
X
Coastal and oceanic
X
Meristic
First dorsal spines
X-XII
—
_
—
—
—
—
Second dorsal spines
and
rays
23-2T
Vl-VIll, 15-26
V, 32-37
V-IX. 23-54
38-46-
47-51-
Vll-lX, 25-39
Anal spines and rays
29-32»
111, 14-20
111.21-24
11-111. 16-35
25-32-
33-37-
11-111, 19-24
Pectoral rays
17-19
18-23
19-23
19-22
16-21
18-21
19-23
Precaudal vertebrae
16-18
10
10
10-12
23-25
7
10-11
Caudal vertebrae
29-31
14-15
15
15-20
34-37
9
14-16
Total vertebrae
46-48
24-25
25
25-32
49-61
43-45
25-26
' Total fin elements.
Larvae and Juveniles
The larvae and/orearly juveniles of approximately 28 species
representing 1 3 genera and all six stromateoid families have
been described (Table 163; see Figs. 333 and 334).
Ahlstrom et al. (1976) considered a young stromateoid to be
a larva and not a juvenile if it (1) had not completely formed
rays in all fins and/or (2) had not initiated development of scales.
Scale formation is seldom encountered on specimens that lack
the full complement of fin rays.
Characters of larvae and early juveniles that Ahlstrom et al.
(1976) found useful in distinguishing among stromateoid fishes
include meristics. sequence of fin formation, morphometries,
pigmentation and skeletal features. Of the strictly ontogenetic
characters, sequence of fin formation and pigmentation pattern
are the most important. The overlap in character states and the
lack of data for certain genera strongly limit the number of
unambiguous characters useful in developing a stromateoid
phylogeny.
Table 163. Stromateoid Species for which Eggs and/or Larvae Have Been Described.
Species
Eggs
Amarsipidae
Centrolophidae
Nomeidae
Tetragonuridae
Ariommidae
Stromateidae
Amarsipus carlsbergi
Schedophilus ovalis
Centrolophus niger
Senolella brama
Seriolella caerulea
Seriolella punctata
Icichthys lockingtoni
Cubiceps baxteri
Cubiceps caeruleus
Cubiceps capensis
Cubiceps gracilis
Cubiceps paradoxus
Cubiceps pauciradiatus
Cubiceps squamiceps
Nomeus gronovii
Psenes arafurensis
Psenes cyanophrys
Psenes maculatus
Psenes pellucidus
Psenes sio
Telragonurus atlanticus
Telragonurus cuvieri
Tetragonurus pacificus
Ariomma regutus
Siromaleus fiatola
Peprilus burli
Peprilus paru
Peprilus simtltimus
Peprilus triacanthus
Pampus chinensis
X?
X?
X
X
X
Ahlstrom et al. (1976)
Padoa (1956b)
Sanzo (1932b) (as C pompilus)
Grimes and Robertson (1981)
Grimes and Robertson (1981)
Grimes and Robertson (1981)
Ahlstrom et al. (1976)
Ahlstrom et al. (1976) (as C. caeruleus. see Butler, 1979)
Ahlstrom et al. (1976) (as C. capensis, see Butler. 1979)
Ahlstrom et al. (1976) (as C. sp. A. see Butler. 1979)
Sparta (1946)
Ahlstrom et al. (1976) (as C sp. B, see Butler. 1979)
Ahlstrom et al. (1976)
Nellen (1973b) (as Psenes whiteleggii. see Ahlstrom et al.,
Ahlstrom et al. (1976)
Ahlstrom et al. (1976)
Ahlstrom et al. (1976)
Ahlstrom et al. (1976)
Ahlstrom et al. (1976)
Ahlstrom et al. (1976)
Grey (1955b), Ahlstrom et al. (1976)
Grey (1955b), Ahlstrom et al. (1976)
Grey (1955b), Ahlstrom et al. (1976)
McKenney (1961) (as Psenes regulus)
Padoa (1956b)
Ditty and Truesdale (1983)
Martin and Drewry (1978), Ditty and Truesdale (1983)
D" Vincent et al. (1980)
Colton and Honey (1963), Martin and Drewry (1978),
Ditty and Truesdale (1983)
Pali (1979)
1976)
' Early juveniles
HORN: STROMATEOIDEI
Table 162. Extended.
623
Cenlrolophidae
Nomeidae
Anommidae
Tetragonundae
Stromateidae
Psenopsis
Cubtceps
Notneus
Psenes
Ariomma
Telragonurus
Siromaieus
Pepnius
Pampus
X
X
X
—
X-XII
Xl-XII
IX-XII
XI-XII
XI-XVIII
_
_
V-VII, 26-
-32
I, 15-27
25-27"
I, 19-30
I, 13-17
10-13
42-56'
II-
-IV, 38-
■49
0-X, 33-50
II-IV, 22-
•27
II
-III, 14-23
II, 24-26
II-
-III, 19-30
II
-III, 13-16
I-II, 8-10
II-
-III, 31-
45
II-
-IV, 35-
■47
0-VII, 39-47
16-23
17-23
21-23
16-22
20-25
14-21
18-25
17-24
24-27
10
12-13
14
12-13
13-14
18-27
18-19
12-15
14-16
15
18-21
27
19-29
17
20-28
24-26
16-22
19-26
25
30-34
41
31-42
30-31
39-54
41-49
29-36
33^1
Behavioral and morphological features of young stromateoids
are potentially informative as taxonomic characters. Certain
ones of these traits appear to be related to the widespread as-
sociation of these fishes with a variety of floating objects in the
ocean. In general, loss of the swimbladder accompanies allo-
metric growth in pectoral fin length and changes in pigmentation
pattern as part of the transition from the juvenile to the adult
stage (Horn, 1975). Stromateoid fishes associated with floating
objects usually have conspicuous blotches or bands of pigment
on their bodies as juveniles then become more uniformly pig-
mented as deeper-living, presumably independent and contin-
uously swimming adult fish. Haedrich (1967) proposed that
banding is protective coloration for the fishes during the period
when they live in the shifting shadows beneath jelly fishes. There
are exceptions to this apparent relationship between pigmen-
tation and behavior. For example, juveniles of Ariomma are
banded yet appear to seldom associate with floating objects
whereas young Telragonurus are uniformly pigmented but, as
Janssen and Harbison (1981) observed, associate intimately with
salps and pyrosomes. The re?ra^o«Mn/5-salp/pyrosome asso-
ciation, however, is different in that the fish are inside rather
than beneath the floating objects. Pigmentation pattern and type
of association are the two ontogenetic characters used in the
phylogenetic analysis (see below).
Fin characters. — Meristic characters (Table 1 62) have been used
widely to distinguish stromateoid taxa especially at the species
level (e.g., Haedrich, 1967; Haedrich and Horn, 1972; Horn,
1970b, 1973; Horn and Haedrich, 1973; Ahlstrom et al., 1976;
Butler, 1979; McDowall, 1982). As in most other perciform
fishes, the pelvic fin (I, 5) and caudal fin (17 principal rays, 15
branched) of stromateoids have stabilized counts (the pelvic fin,
however, is absent in three stromateoid genera). The number
of secondary caudal rays, although exhibiting intraspecific vari-
ation, can be an important taxonomic character among species
within a genus (Ahlstrom et al., 1976). The dorsal fin of stro-
mateoids may be continuous or divided into two fins. This trait
is used as a generic character in the present paper (Tables 164,
1 65). It is not always possible to distinguish between spines and
rays in those species with a continuous dorsal fin (see Table
162). The complement of anal fin rays in stromateoids is pre-
ceded by 0 to 7 anal spines with most species having 2 or 3
spines. The number of pectoral fin rays varies from 14 to 27
among stromateoids, but the overlap among species limits its
use as a taxonomic character.
Two different sequences of fin formation occur in oceanic
stromateoids depending primarily on whether the pelvic fins
form early (before the other fins) or whether they form late.
Ahlstrom et al. (1976) found that the pelvics are first to form
in Amarsipus, Psenes and probably also Nomeus whereas they
are last to form in Cubiceps, Icichthys and Telragonurus. Fahay
(1983) reported that the pelvics are also last to form in Centro-
lophus. These ontogenetic patterns have potential significance
as taxonomic characters; however, they must be described for
other genera before they can contribute to an understanding of
stromateoid relationships.
Morphometries. — SXrom?i\eo\ds vary substantially in their mor-
phologies, especially body depth, but show no abrupt meta-
morphic changes in the transition from the larval to the juvenile
to the adult stage. Allometric growth is common in these fishes
and complicates the use of morphometries as taxonomic char-
acters. Taxa at similar stages of development must be compared
if morphometric characters are to have validity. Ahlstrom et al.
(1976) used morphometries in distinguishing among species in
genera such as Schedophilus and Psenes. Because of allometry
and the less than complete information on different develop-
ment stages of several genera, morphometric characters were
not used in the phylogenetic analysis of stromateoid genera (see
below).
Skeletal characters. — M\\s,Xrom et al. (1976) in their study of
the early life history stages of oceanic stromateoids found the
following skeletal characters to be of particular relevance: (1)
total number of vertebrae, (2) co-occurrence of a pair of pleural
ribs and a haemal spine on each of one or more caudal vertebrae,
(3) separation of vertebrae into precaudal and caudal groups,
(4) position of anal fin pterygiophores in relation to haemal
spines, (5) number and position of dorsal fin pterygiophores and
predorsal bones in relation to neural spines, and, (6) the number
of supporting bones of the caudal fin. While not strictly onto-
genetic in nature, these characters are most readily discerned
from examiniation of cleared and stained larvae and early ju-
veniles.
Of the above characters, only the number of predorsal bones
and the number of hypurals were used in the phylogenetic anal-
^<<Z2:^?^'^-^^ '^'"'^y^.s
B
Fig. 334. Examples of stromateoid larvae and early juveniles. (A) Cuhiceps pauciradiatus (Nomeidae), 17.5 mm early juvenile; (B) Psenes
cyanophn's (Nomeidae), 19.1 mm early juvenile; (C) Tetragonurus atlanticus (Tetragonuridae), 17.2 mm postfiexion larva; (D) Ariomma sp.
(Ariommidae), 14.4 mm early juvenile. Gulf of Mexico; (E) Pepnius similtimus (Stromaleidae), 10.8 mm postflexion larva. A-C from Ahlstrom
et al. (1976), D drawn by Betsy Washington. E from D'Vincent et al. (1980).
HORN: STROMATEOIDEI
625
Table 164. Characters and Character States Used in the Phylogenetic Analysis of Stromateoid Genera.
Character stale codes
Character
0
1
2
3
1 . Number of rows of premaxillary teeth
1
2
2. Number of rows of dentary teeth
1
2
3. Pharyngeal sac
absent
present
4. Shape of pharyngeal sac
height > length
height = length
height < length
5. Arrangement of papillae in
10-20 bands
5-7 bands
not in bands
pharyngeal sac
6. Papillae on upper portion of
present
absent
pharyngeal sac
7. Position of papillae in pharyngeal
not on stalks
short stalks, teeth
long stalks, teeth
sac
on end
along stalk
8. Shape of papillae base in pharyngeal
sac
9. Condition of maxilla
round
stellate
mobile
fixed
10. Supramaxillary bone
present
absent
1 1. Lacrimal bone
prominent
reduced
highly reduced
12. Relationship of gills to isthmus
free
united
13. Pseudobranch
present
absent
14. Scale type
cycloid
ctenoid
15. Opercular scalation
present
absent
16. Preopercular scalation
present
absent
17. Prominent preopercular spines
absent
present
18. Number of branchiostegal rays
7
6
5
19. Pelvic bone and fin
fin present
fin absent, bone
fin absent.
fin absent, highly
with spine
reduced bone
reduced bone
20. Number of predorsal bones
3
2-3-
0
7-12
21. Number of dorsal fins
1
2
22. Keels on caudal peduncle
absent
present
23. Number of hypurals
6
3-5
2
24. Procurrent spur
present
reduced
absent
25. Ray base preceding procurrent spur
shortened
slightly shortened
not shortened
26. Juvenile pigmentation
uniform
patterned
27. Primary juvenile association
independent
floating objects
inanimate floating
objects
animate floating
objects
■ This character slate overlaps adjacent stale but occurs in only one taxon (Girellidae).
ysis of stromateoid genera (Tables 1 64, 165; Fig. 335). The other
characters were not used because they are not known for all
genera or, if known, their values overlap and, therefore, cannot
be coded without ambiguity.
The pharyngeal sac as a skeletal feature was a rich source of
characters in developing the phylogenetic hypothesis for stro-
mateoid genera. Five characters were used ranging from the
shape of the sac to the arrangement, location and position of
papillae within the sac (Tables 164, 165; Fig. 335).
Pigmentation. — Differences in pigmentation are mainly of value
in distinguishing species within genera for which the larval and
juvenile stages are relatively well known. Ahlstrom et al. ( 1 976)
used pigmentation patterns to demonstrate differences among
species of oceanic stromateoid genera (see Figs. 333 and 334).
Adults tend to be more uniform in pigmentation and, hence,
offer fewer apparent taxonomic characters.
Stromateoids vary both in the density and in the pattern of
their pigmentation. As larvae and early juveniles, some species,
e.g., Amarsipus carlshergi (Fig. 333A) are sparsely pigmented
whereas others are more heavily pigmented, e.g., Icichthys lock-
inglom (Fig. 333C). Certain larvae and juveniles are rather uni-
formly pigmented, e.g., Ciibiceps paiuiradiatus {Fig. 334A), 7et-
ragonunis atlanticiis (Fig. 334C) and Pepriliis simillimus (Fig.
334E) while others have their pigment concentrated into bands
or blotches, e.g., Schcdophilus huttoni (Fig. 333B). Numeus
gronovii (¥{%. 333D). Psenes cyanophrys (Fig. 334B)and /lr/o«;-
ma sp. (Fig. 334D). Ahlstrom et al. (1976) used various detailed
patterns to distinguish the larvae, especially, and early juveniles
of species within certain stromateoid genera. In the present study,
uniform vs patterned pigmentation was the only pigmentation
character available that could be coded unambiguously for all
stromateoid genera (Tables 164, 165).
Relationships
Relationships within the Stromateoidei
Haedrich's (1967) analysis continues to be the major system-
atic work on stromateoid fishes. He recognized five families and
two main lineages in the stromateoids. One lineage is composed
of the Centrolophidae and its derivative, the Stromateidae. The
other, a less compact assemblage, is comprised of the Nomeidae
and its two derivatives, the Ariommidae and the Tetragonuri-
dae. The Centrolophidae and the Nomeidae contain the basal
stocks with the centrolophids having the more primitive mem-
bers. Haedrich (1967) considered members of the centrolophid
genus Hyperoglyphe to be the most generalized fishes in the
suborder and probably not unlike the ancestral form. He viewed
the Stromateidae as the current zenith of stromateoid evolution
with Pampus as the most advanced stromateid genus. In his
interpretation of stromateoid relationships. Haedrich (1967)
recognized trends in the evolution of several characters includ-
626
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 165. Matrix of Character State Codes (see Table 164 and Fig. 335 for the stromateoid genera and for the three perciform families
used as out-groups in the phylogenetic analysis). Dashes indicate characters that are inapplicable. Question marks indicate unknown character
states and were designated as "missing observations" in the analysis.
1 . Number of rows of premaxillary teeth
2. Number of rows of dentary teeth
3. Pharyngeal sac
4. Shape of pharyngeal sac
5. Arrangement of papillae in pharyngeal sac
6. Papillae on upper portion of pharyngeal sac
7. Position of papillae in pharyngeal sac
8. Shape of papillae base in pharyngeal sac
9. Condition of maxilla
10. Supramaxillary bone
1 1. Lacrimal bone
12. Relationship of gills to isthmus
13. Pseudobranch
14. Scale type
15. Opercular scalation
16. Preopercular scalation
17. Prominent preopercular spines
1 8. Number of branchiostegal rays
19. Pelvic bone and fin
20. Number of predorsal bones
21. Number of dorsal fins
22. Keels on caudal peduncle
23. Number of hypurals
24. Procurrent spur
25. Ray base preceding procurrent spur
26. Juvenile pigmentation
27. Primary juvenile association
Siromateoid genera
Amarsipus
Hyperoglyphe
Centrolophus
Schedophtlus
Icicfilhys
Tuhbm
SerioleUa
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
—
0
0
0
0
0
0
—
0
0
0
0
0
0
—
0
0
0
0
0
0
—
0
0
0
0
0
0
—
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
1
1
1
0
0
1
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
?
0
2
?
0
0
0
7
0
2
?
2
0
1
1
1
1
1
1
0
2
3
3
3
3
3
ing body size and shape, fin pattern, presence or absence of
palatal dentition, shape of the papillae in the phai7ngeal sacs
and the number of branchiostegal rays, vertebrae and epural
plus hypural elements in the caudal skeleton.
Haedrich ( 1 969) in describing the Amarsipidae added a sixth
family to the stromateoid suborder. This family exhibits a mix-
ture of primitive and derived characters. It lacks pharyngeal
sacs, but the pharyngeal teeth are extraordinarily developed and
may perform a shredding function analogous to the sacs of other
stromateoids. Haedrich (1967) argued that possession of a per-
ciform caudal skeleton, uniserial jaw teeth, an expanded lacri-
mal bone, an inflated and protruding top of the head, an exten-
sive subdermal canal system and a bony bridge over the anterior
vertical canal of the ear provides the basis for placement of
Amarsipus in the suborder Stromateoidei. He considered the
new family to be distantly allied with the Nomeidae.
In the present study, a phylogeny of the 1 6 stromateoid genera
was constructed using a set of 27 characters (Tables 164, 165)
that could be coded for all genera with little or no overlap and
ambiguity. Initially, a larger number of prospective characters
(~65) were identified and evaluated. Comparison of stroma-
teoids with presumed out-groups helped to generate characters
and to establish polarity in the transformation series. Charac-
ters, however, were omitted if they could not be quantified or
if insufficient information was available to characterize every
taxon. In a few cases, character states were coded as "missing
observations" if three or fewer genera required this coding and
the characters were judged important in resolving relationships
between the other genera.
Three closely related perciform families— Girellidae, Ky-
phosidae and Scorpididae— were used as potential out-groups
in the analysis (see below for rationale). Although these three
taxa are frequently classified as subfamilies of the Kyphosidae
(e.g.. Nelson, 1976), G. D. Johnson (pers. comm.) considers
them to be distinct families.
The analysis was performed using the PHYSYS package which
differentiates taxa based on the presence of shared derived char-
acters (synapomorphies). Several phylogenetic trees were gen-
erated from the genus-character data matrix using the Wagner
distance algorithm (see Farris, 1970; Wiley, 1981). These trees
were diagnosed to identify the origin of each apomorphy and
to examine character reversals and convergences. Transfor-
mation Series Analysis was performed on the data to verify
polarities developed through out-group comparison (see Mick-
evich, 1982) and to resolve nonlinear series. Further optimizing
produced the most parsimonious tree from the data matrix (Fig.
335).
This phylogenetic tree (Fig. 335) was basically similar, with
certain exceptions, to that proposed by Haedrich (1967). //i'-
peroglyphe emerged as the most plesiomorphic stromateoid tax-
on possessing a pharyngeal sac and Pampus as the genus with
the greatest number of apomorphies. Arioinma also ranked as
a highly derived genus in the suborder. Despite its several syn-
apomorphies with advanced stromateoid genera, Amarsipus
emerged as the sister taxon of all other stromateoid genera pri-
marily because it lacks a pharyngeal sac. The major differences
between the present analysis and Haedrich's interpretation lie
with the relationships of Arioinma and Tctragonurus to other
stromateoids and with the family limits of the suborder. Based
on the cladogram, Tctragonurus and Ariomma are more closely
HORN: STROMATEOIDEI
Table 165. Extended.
627
Slromateoid genera
Perciform families
Psenopsts
Cubiceps
Nomeiis
Psenes
Anomma
Tetragonurus
Stromateiis
Pepnliis
Pampus
Kyphosidae
Scorpididae
Girellidae
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
2
2
1
1
1
0
1
1
1
2
2
2
2
2
0
0
0
0
1
0
0
0
0
0
1
1
1
2
1
2
2
2
0
1
1
1
0
0
1
1
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
2
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
1
0
0
0
1
1
1
I
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
2
0
0
1
0
0
0
0
0
0
3
1
2
0
0
0
0
0
0
0
0
2
0
0
0
0
0
1
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
1
1
1
2
I
1
1
1
1
1
1
?
1
0
0
2
2
0
0
0
0
0
0
7
1
1
2
2
2
2
2
2
0
0
2
0
0
1
1
1
0
1
0
0
1
0
0
3
3
3
1
0
3
3
3
3
1
0
1
related to the stromateids than to the nomeids as proposed by
Haedrich (1967). Of the six families recognized by Haedrich
and Horn (1972), only the Amarsipidae, Nomeidae and Stro-
mateidae appear to be monophyletic based on the generic re-
lationships expressed in the cladogram. The Centrolophidae, on
the other hand, lacks a synapomorphy and there is no indication
that Tetragonuridae and Ariommidae should be considered dis-
tinct families. Additional characters, however, should be ex-
amined before a change in classification is proposed.
The Scorpididae and Girellidae were part of a trichotomy at
the base of the tree and together formed the plesiomorphic out-
group cluster in the analysis.
Character diagnosis showed that reversals in character trans-
formation occurred most frequently (> 3 taxa or stems/character
state) for scale type (no. 14) and juvenile pigmentation (no. 26).
In the same diagnosis, character convergences occurred most
often for the supramaxillary bone (no. 10), opercular scalation
(no. 15), number of branchiostegal rays (no. 18), number of
dorsal fins (no. 21)andjuvenileassociations(no. 27). Atnarsipus
was involved in all five of these cases of apparent convergence,
an indication of its uncertain phylogenetic position.
The tree remains incompletely resolved with three polychot-
omies (Fig. 335). In addition to the trichotomy at the base of
the tree, the other two nodes with multiple branches involve
centrolophid genera. Lack of full resolution in this region of the
tree indicates that further work is needed to clarify the inter-
generic relationships of the Centrolophidae. Extending the anal-
ysis to the species level would provide greater resolution.
Strictly larval or juvenile characters have contributed little to
the broad understanding of stromaleoid intergeneric relation-
ships as perceived by Haedrich (1967) or as analyzed in the
present study. Elimination of the two juvenile characters (pig-
mentation and associations) from the present analysis resulted
in a tree virtually identical to that with them included (Fig. 335).
The study of the early life history stages of pelagic stromateoids
by Ahlstrom et al. (1976), however, is a major contribution to
stromateoid systematics especially in developing an approach
that can potentially expand to all taxa in the suborder. Their
use of ontogenetic characters was important at the species level
and particularly valuable in distinguishing the species and gen-
era of nomeids.
Characters employed by Ahlstrom et al. (1976) that hold
promise for resolving relationships among stromateoids in gen-
eral include ( 1 ) sequence of fin formation, (2) arrangement of
anal fin pterygiophores in relation to haemal spines, (3) head
armature and (4) pigmentation patterns. The caudal fin complex,
while not representing a strictly ontogenetic suite of features,
also appears likely to provide characters if a full spectrum of
cleared and stained larvae are carefully examined. Finally, the
various types of associations juvenile stromateoids hold with
floating objects may be more specific than generally thought and
could become a rich source of characters.
Relationships of the Stromateoidei to
other groups
Haedrich ( 1 967) in his review of stromateoid systematics pro-
posed that the group arose from within a relatively undiffer-
entiated assemblage of perciform families including the Arri-
628
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Out-Groups Amarsipidae
C e n t r o I o p h i d a e
Ariommidae
Tetragonuridae
^®
» . .o^ •;;?>' -c*'
.<f
/ / / / / / / / r /
Nomeidae
S t r 0 m a t e i d a e
^ r /' J"' ./ ./ y y / '^
# / /" /" ^i-- v^" ,.'• /■ ,»"'
-27
- 18
_
-10
_
7
- 10
-20
^24 -
- 14
- 14 -
- 6
- 26
H
- IS
-
- 20
-
- 11
- 15
- 24
- 26 -
-27 -
- 20 -
- 7
- 25
-
- 15
- 18
- 21
- 26
- 27
- 26
- 26 -
- 15
- 16
- 23
- 27
-
- 19
-
-25
- 4
- 22
21
24
- 7
8
= "l9
--8
_
- 4
--21
- 5
4-
- 5
7
'z':
:'°i8
- 23
-
- 16
S t romal e
- 25
Tetragonuridae
Ariommidae
■ 27
\ Nomeidae /
- 3
- 27
Amarsipidae /
14
--16
- - 23
Centrolophidae /
/ /
_
_ 1
\ j ^^
I
- 2
- 26
'--. /
H
aec
rich'
s (1967)
Tree
Fig. 335. Phylogenetic hypothesis of relationships among the genera of stromateoid fishes. The perciform famihes Scorpididae, Girellidae and
Kyphosidae are out-groups. Each number represents a character and each honzontal bar represents a character state indicating a synapomorphy
or autapomorphy. Character transformation series 1-27 are shown in Table 165. Relationships of the stromateoid families as perceived by
Haedrich (1967) are shown, with addition of the Amarsipidae, in the lower right. Limits of the six stromateoid families as recognized by Haedrich
and Horn (1972) are shown at the top of the cladogram.
pididae, Girellidae, Kuhliidae, Kyphosidae, Pomatomidae,
ScoiTJididae and Theraponidae. All are Pattern- 10 teleosts in
terms of the ramus lateralis accessorius (a facial nerve complex)
and have a bony bridge over the anterior vertical canal of the
ear. Of these families, the Kyphosidae bear a strong resemblance
to the Centrolophidae, the most primitive stromateoids. Hae-
drich (1967) implicitly favored the hypothesis that the kyphos-
ids are the closest perciform relatives of the most primitive
stromateoids pointing out that both have 10+15 vertebrae, a
caudal skeleton with six hypural and three epural elements (ac-
tually five hypurals in kyphosids), a perforate ceratohyal, seven
branchiostegals, an expanded lacrimal bone and scaled fin bases.
The present analysis (Fig. 335) supports this hypothesis.
Ontogenetic characters have not been used previously in ana-
lyzing the relationships of stromateoids to other perciform groups.
Use of the two juvenile characters in the present analysis did
not alter the phylogeny based on adult characters. Although they
were not examined in the present study, kyphosid larvae arc
more generalized than girellid or scorpidid larvae and, therefore,
more similar to stromateoid larvae (G. D. Johnson, pers. com-
mun.). The generalized nature of stromateoid larvae suggests
that their characters will continue to be most useful in distin-
guishing species (e.g.. Ditty and Truesdale, 1983) and less valu-
able at higher levels of classification.
Department of Biological Science, California State
University, Fullerton, California 92634.
Gobiesociformes: Development and Relationships
L. G. Allen
THE Gobiesociformes includes three families, the Gobiesoci-
dae, Callionymidae and Draconettidae, according to Gos-
line (1970) and Nelson (1976). Members of this order are pri-
marily marine bottom -dwellers in shallow-waters and occur
worldwide in tropical and temperate seas. Distinguishing char-
acteristics of the order include: a scaleless head and body; 5-7
branchiostegal rays; no circumorbital bones behind the lacrimal;
articular processes of the premaxillae either fused with ascending
process or absent; pelvic fin base well in advance of pectoral
fin; no swim bladder (in adults) (Nelson, 1976). The order con-
tains about 54 genera with 246 species in the three families
(Bnggs, 1955; Nakabo, 1982a. b).
Briggs' (1955) review of the Gobiesocidae remains as the most
thorough treatment of this family to date. Revisions of both the
Callionymidae and the Draconettidae have recently been pub-
lished by Nakabo (1982a, b). Hypotheses of systematic rela-
tionships within the families are based entirely on adult char-
acteristics.
The Callionymidae (dragonets) is a large and diverse group
within the Gobiesociformes. The ontogeny and systematics of
the Callionymidae is presented in this volume by E. D. Houde.
The smallest family of the order, the Draconettidae, consists
of two genera and seven species (Nakabo, 1982a). Draconettids
are small demersal fishes inhabiting sand-mud bottoms along
the edge of the continental shelf or on seamounts. They occur
widely in tropical and temperate waters of the world except the
eastern Pacific. Adult draconettids resemble callionymids which
lead one author (Davis, 1 966) to include the draconettids within
the Callionymidae. Gosline (1970) and Nakabo (1982a) dis-
agreed with this inclusion.
Table 166. Egg Characteristics of 18 Species of Gobiesocids for which Larvae Are Known Organized into Subfamilies after Briggs
(1955).
Eggc
laraclenstics
Size {mm)
Number eggs
Oil
Species/ Reference
SublamiK
Shape
(long axis)
in mass
globules
Color
Where laid
Conidens lalicephatus
Trachelochisminae
flat, ellipsoid
1.28-1.38
77-109
+
green-
under rock
(Shiogaki and Dolsu. 197 Id)
orange
Trachclochismus melobesia
Trachelochisminae
flat, oval
.?= 1.65
3—300
10-100
red/pink
under rock
(Ruck. 1971)
Trachclochismus pinnutalus
Trachelochisminae
oval
.?= 1.81
198-1,500
1-6
yellow-
under rock
(Ruck, 1973b)
red
Lepadogaster lepadogaster
Lepadogastrinae
flat, oval
1.8-1.9
200-250
1
yellow/
under rock
(Guitel, 1888; Russell, 1976)
amber
Lepadogaster candolei
Lepadogastrinae
flat, oval
1.2
—
1
yellow
under rock
(Guitel, 1888; Russell, 1976)
Apletodon rnicrocephalus
Lepadogastrinae
—
—
—
1 to
—
kelp stems
(Guitel, 1888; Russell, 1976)
several
Diptecogaster biinaculata
Lepadogastrinae
flat, oval
1.37-1.54
—
1
—
empty shells
(Guitel, 1888; Russell, 1976)
,
Diplocrcpis pumccus
Diplocrepinae
spherical
1.80
< 2,400
20-30
purple
under rock
(Ruck, 1973b)
Gastroscvphus hectoris
Diplocrepinae
—
—
—
—
—
—
(Ruck, 1976)
Gastrocvalhus gracilis
Diplocrepinae
—
—
—
—
—
—
(Ruck, 1976)
Acrvlops bervllmus
Gobiesocinae
oval
1.1
2-40
2.5
green-
Thallasia
(Gould, 1965)
yellow
blades
Gobiesox maeandricus
Gobiesocinae
oval
1.68-1.92
—
—
—
under rock
(Allen and llg. 1983)
Gobiesox rhessodon
Gobiesocinae
oval
—
1 50-200
—
orange
under rock
(Allen, 1979)
Gobiesox slrumosus
Gobiesocinae
elongate oval
0.75-0.94
650-2,500
70-80
—
empty shells
(Runyan. 1961; Dovel, 1963)
Rimicola muscarum
Gobiesocinae
—
—
—
—
—
kelp blades
(Allen, 1979)
Lepadichlhys frenatus
Diademichthyinae
flat, ellipsoid
1.31-1.36
240-301
1-6
—
shell
(Shiogaki and Dotsu, 1971b, c)
Aspasma minima
Aspasminae
ellipsoid
1.25-1.35
140-619
20
yellow
under rock
(Shiogaki and Dotsu, 1971a)
Aspasmichlhys ciconiae
Aspasminae
—
—
—
—
—
—
(Shiogaki and Dotsu, 1972d)
629
630
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 167. Larval Characterlstics of 18 Species of Gobiesocids for which Larvae are Known Organized into Slibfamilies after Briocss
(1955). See Fig. 338 for abbreviations for regions of pigmentation (coding of pigment patterns depended on illustrations in most species. * denotes
counts on older postflexion larvae).
Pigmentation (left side)
Species/Reference
Subfamily
DH
DT
LT
1X3
Trachelochisminae
5
6-10
3-7
11-15
Trachelochisminae
0
0
0
26-35
Trachelochisminae
0
0
0
14
Lepadogastrinae
8
23-24
43
11
Lepadogastrinae
4
0
5-6
0
Lepadogastrinae
4
15
36
13
Lepadogastrinae
0-4
13-25
32-62
0
Diplocrepinae
0
0
30-50
19-23
Diplocrepmae
0
0
2-3
19-23
Diplocrepinae
0
0-2
7-10
15-25
Gobiesocinae
?
+
+
+
Gobiesocinae
0
0
0
9-15
Gobiesocinae
0-5
0
8-17
13-15
Gobiesocinae
0-10
10-15
0-24'
9-12
Gobiesocinae
7
0
10-50
21
Diademichthyinae
5-12
15-21
70-105
12-19
Aspasminae
0
0*-13
10-13
7-17
Aspasminae
5
16-17
24-31
6-7
Conidens lalicephalus (Shiogaki and Dotsu, 197 Id)
Trachelocfusmus inelobesia (Ruck. 1971)
Trachelochismus pinnulatus (Ruck, 1973b)
Lepadogasler lepadogasier (G\i\Xc\. 1888; Russell, 1976)
Lepadogasler candolei {Guitel, 1888; Russell, 1976)
Apletodon microcephalus (G\i\xe\. 1888; Russell, 1976)
Diplecogasler bimaculala (Guileh 1888; Russell, 1976)
Diplocrepis pumceus (Ruck, 1973b)
Gaslroscyphus hecloris (Ruck, 1976)
Gastrocyalhus gracilis (Ruck, 1976)
Acrytops beryllinus (Gould, 1965)
Gobiesox maeandricus (Allen and llg, 1983)
Gobiesox rhessodon (Allen, 1979)
Gobiesox slrumosus (Kuny&n, 1961; Dovel, 1963)
Rimicola muscarum (Allen, 1979)
Lepadichthys frenatus (Shiogaki and Dolsu, 1971b, c)
Aspasma minima (Shiogaki and Dotsu, 1971a)
Aspasmichlhys ciconiae (Shiogaki and Dotsu, 1972d)
The early life history stages of draconettids are unknown at
this time. Therefore, the Draconettidae will receive little further
attention in this paper.
The Gobiesocidae (clingfishes) is a diverse group of primarily
shallow water or intertidal marine (although a few species are
freshwater) fishes consisting of about 33 genera and 100 species.
Clingfishes occur along tropical and temperate shores in the
Atlantic, Indian and Pacific Oceans. Distinguishing character-
istics of gobiesocids include: pelvic fins modified into a thoracic
suction disc; pelvic fin with one small modified spine and four
or five soft rays; single dorsal fin without spines; no basibran-
chials; vertebrae 25-54 (or 78 if the genus Alabes is included,
see below); lateral line confined to head; two postcleithra; hy-
purals fused into a single plate. Most species are small (normally
<70 mm), but a few attain relatively large size (up to 300 mm)
(Nelson, 1976). Eight subfamilies are recognized (Briggs, 1955).
Springer and Fraser (1976) synonymized the family Cheilo-
branchidae (=Alabetidae) with the Gobiesocidae based on shared
specializations particularly of the structure of the joint between
the supracleithrum and cleithrum. If valid, this synonymy adds
one more genus (Alabes) and four species to the Gobiesocidae.
Development
Eggs
Spawning occurs in rocky intertidal or subtidal areas. Eggs
are demersal and are attached to the underside of rocks or shells
or on kelp blades. The adults (usually the male) guard the eggs
during development. Eggs are ovate to ellipsoidal in shape and
range from about 0.7 to 1.9 mm in greatest dimension (Table
166). The monolayered egg masses usually contain between 100
and 600 eggs with some reports of up to 2,500 eggs in a patch
(Gobiesox strumosus) (Table 166). The initial coloration of the
eggs ranges between purple and green (with pink, yellow and
orange predominating). Eggs contain anywhere from one to 100
oil globules depending on species and stage of development. As
development proceeds, oil globules tend to coalesce ultimately
into one.
Larvae
Larvae of 18 species of gobiesocids have been described in
varying detail (Table 167). Larval series are available for 10 of
these species. Larvae are well developed at hatching and possess
functional jaws, fully pigmented eyes, body pigment similar to
that of later larval stages and a small (sometimes bilobed) yolk
sac in most species which is probably absorbed soon after hatch-
ing. Size at hatching ranges from 2.4 mm in Gobiesox strumosus
to 6.8 mm in G. maeandricus and appears to be related to egg
size and maximum size of the adults (Table 168). Larvae are
cylindrical and somewhat laterally compressed becoming more
robust with growth. All clingfish larvae have long, underslung
guts which usually extend beyond the midline of the body (preanal
length 50-70% SL) in both pre- and postflexion larvae. Size at
notochord flexion was difficult to determine from most of the
larval descriptions, but generally ranged between 5.0 and 8.0
mm depending on the species (Table 167). Gobiesocid larvae
have well-developed swimbladders in the early stages of devel-
opment which are located in the dorso-anterior portion of the
peritoneal cavity. In several species the swimbladder is hidden
by the heavy pigmentation on the dorsum of the gut. No sudden
change occurs at settlement; rather, larvae gradually attain ju-
venile characteristics. Size at settlement is, therefore, difficult
to determine. Juvenile characteristics are attained at a wide
range of sizes (6.3-13.0 mm) with a trend toward larger species
"settling" at larger sizes (Tables 167 and 168). Presumably, the
loss of the swimbladder occurs during settlement.
Most gobiesocid larvae are heavily pigmented. Furthermore,
the numbers and patterns of the large, stellate melanophores on
the body are species specific, and are invaluable in the identi-
fication of species (Figs. 336 and 337). Melanophores occur
primarily in the seven regions designated in Fig. 338. Larvae of
ALLEN: GOBIESOCIFORMES
Table 167. Extended.
631
Pigmentation (left side)
Myomere
counl
Size (mtn) at
_ Preanal
lenglh/SL (%)
Hatching
Notochord
flexion
Onset of Pj
develop
Settling
LG
VG
PV
Preflexion
Postflexion
0
11-14
4-5
26-31
3.4
4.5-5.5
4.5
6.3
61.4
72.4
0
0
5-9
31-36
4.8-5.5
6.3-7.0
6.5
7.8?
62.9
64.9
0
0
0
31
5.3-6.1
_
—
59.6
16
13
12
—
5.1
—
_
_
62.1
_
12
1
9
—
5.0-6.0
—
6.5?
—
57.2
—
20
0
2
—
—
_
_
7.0?
59.9
15-20
19
0
—
3.0-4.3
4.5-6.5
8.0
_
63.5
72.4
9-26
0
5-6
—
5.0-6.0
6.2-7.7
7.7
65.7
65.0
0
3
12-16
—
—
—
_
_
61.8
60.2
4-7
4
0-1
—
—
5.5-6.9
—
—
60.7
72.3
+
+
9
_
2.6
0
0
9-16
32-35
5.6-6.8
6.7-7.0
7.3
13.0
50.0
63.6
0
0
4-7
24-29
3.9-4.1
5.5-6.9
5.5
-10.0
58.0
68.0
0
0-1
4-8
28-29
2.4-3.4
4.7-6.5
6.6
-12.0
61.4
72.9
0
0
0
—
4.0
—
—
55.3
0-4
0
9-14
36-37
4.2
6.0
6.0
9.9
58.8
79.0
0-12
2-4
6-12
35
3.6-4.0
5.5-6.8
4.5
6.7-7.4
67.3
76.7
6-21
14-15
0
—
-
-
-
-
62.7
75.5
each species exhibit a unique distribution of melanophores with-
in and among these regions (Table 167). The distribution of
melanophores within regions can be coded. For example, Con-
idcns laticephalus has a pigment pattern which can be designated
as the followmg: DH5, DT6-10, LT3-7, DGl 1-15. LGO, VGl 1-
14, PV4-5. Trachelochismus melohesia by the same process is
designated as: DHO, DTO, LTO, DG26-35, LGO, VGO, DV5-
9. If adopted, this system of coding pigment patterns will serve
two purposes. It will greatly aid identification of clingfish larvae
and will also lead to more comparable descriptions of gobiesocid
larvae in the future.
Pigment patterns do not appear to be related to phylogenetic
hypotheses based on adult characteristics. In virtually all known
cases closely related species (subfamilial levels) tend to have
noticeably different patterns and often range from heavily to
lightly pigmented (Table 167). Within the Trachelochisminae,
Conidcns laticephalus is heavily pigmented while both Tra-
chelochismus melohesia and T. pinnulala are lightly pigmented.
The same pattern is exhibited in all other subfamilies (Table
167) especially the Lepadogastrinae and Gobiesocinae (partic-
ularly in the genus Gobiesox). Members of subfamilies often
overlap in their distributions (Briggs, 1955). Diverse pigment
patterns among closely related, sympatric clingfish larvae may
well represent ecotypic variation. Heavily pigmented larvae often
live in surface waters where the pigmentation may protect them
against solar radiation or serve as protective coloration (Moser,
1 98 1 ). Less pigmentation may indicate that the larvae normally
occur deeper in the water column where irradiance does not
present problems for development.
Only a few published descriptions included myomere counts.
Those accounts which did revealed a range from 24 up to 37
(Table 167). The number of myomeres appears to have great
diagnostic value in some cases when used in conjunction with
pigment, i.e., among the species of Gobiesox (Table 167). The
lack of myomere count data among the described gobiesocid
larvae may, in part, be due to the difficulty in countmg caused
by heavy trunk pigmentation. Nonetheless, it is unfortunate that
this important character has not received greater attention es-
pecially since vertebral counts are not available for many species.
Adult characteristics which are valuable for identifying older
larvae are also included in Table 168.
The most distinctive characteristic of clingfishes is the suction
disc which is supported by the pelvic fins and distal postcleithra
of the pectoral girdle. The two types of discs are found in go-
biesocids. The "double" disc has a small, posterior disc with a
free anterior margin separating it from an anterior disc. In the
"single" disc the anterior and posterior portions are coalesced
into one continuous structure (Briggs, 1955). The onset of suc-
tion disc development occurs fairly early in larval development
(ranges from 4.5 to 8.0 mm SL) and appears to be closely allied
to time of notochord flexion in most species (Table 167). Disc
development does not appear to differ appreciably between sin-
gle and double disc types except that in the single type a con-
sistent connection remains between the anterior and posterior
elements throughout development (Fig. 339). The completion
of the suction disc is undoubtedly critical in late larval stages.
Settlement seems unlikely to occur without a functional disc.
Specialized glandular tissues appear on the body surface and
out onto the finfolds in several species of gobiesocids (Shiogaki
and Dotsu, 1971b; Allen and Ilg, 1983). Although these struc-
tures are not specifically mentioned in other descriptions, illus-
trations of larvae from some of these studies include structures
in the finfolds which may be these same glandular tissues. Fur-
ther studies are needed to ascertain the extent of this special-
ization within the Gobiesocidae and the possible function of
these tissues.
Relationships
The systematic relationships among the Gobiesocidae were
addressed, as previously mentioned, by Briggs (1955). His eight
subfamilies reflected both morphological similarities and zoo-
geographic distributions (subfamilies occupy fairly distinct re-
B
ALLEN: GOBIESOCIFORMES 633
Table 168. Adult Characteristics for 18 Species of Gobiesocids for which Larvae are Known Arranged by Subfamily (Briggs, 1955).
No.
No.
No
No.
Max.
dorsal
anal
pectoral
caudal
No.
length
Species/Reference
Subfamily
rays
rays
rays
rays
vertebrae
(mm SL)
General distribution
Comdens taticephalus
Trachelochisminae
7-9
5-7
19-20
11-13
28
33.0
Southern
(Shiogaki and Dotsu, 197 Id)
Japan
Tracheloch is mus melobesia
Trachelochisminae
9-11
7-8
22-24
12
—
30.0
New Zealand
(Ruck, 1971)
Trachetochismus pmnulalus
Trachelochisminae
7-9
5-7
24-26
11-12
_
71.2
New Zealand
(Ruck, 1973b)
Lepadogaster lepadogaster
Lepadogastrinae
16-19
9-11
20-23
12-13
—
65.0
NE Atlantic/
(Guild, 1888; Russell, 1976)
Mediterranean
Lepadogaster candolei
Lepadogastrinae
13-16
9-11
26-29
10-13
—
75.0
NE Atlantic/
(Guitel, 1888; Russell. 1976)
Mediterranean
Aptetodon microcephalus
Lepadogastrinae
5-6
5-7
21-24
10-11
—
41.6
NE Atlantic/
(Guitel, 1888; Russell, 1976)
Mediterranean
Dtplecogaster himaculata
Lepadogastrinae
5-7
4-6
21-25
9-10
—
41.0
NE Atlantic/
(Guitel, 1888; Russell, 1976)
Mediterranean
Diplocrepis pumceus
Diplocrepinae
10-11
4-5
23-24
10
—
100.0
New Zealand
(Ruck, 1973b)
Gaslroscyphus hectons
Diplocrepinae
6-8
6-7
20-22
8
—
43.6
New Zealand
(Ruck, 1976)
Gastrocvathus gracilis
Diplocrepinae
5-6
5-7
18-19
8
_
31.0?
New Zealand
(Ruck, 1976)
Acrylops hervtlinus
Gobiesocinae
5-7
5-7
19-23
10
—
20.0
NW Atlantic
(Gould, 1965)
Gobiesox maeandncus
Gobiesocinae
14-16
13-15
21-23
11-13
32-34
114.0
NE Pacific
(Allen and llg, 1983)
Gobiesox rhessodon
Gobiesocinae
12-14
11-12
18-21
11-12
28-29
39.3
NE Pacific
(Allen, 1979)
Gobiesox slrumosus
Gobiesocinae
10-13
9-11
22-26
11-13
25-27
69.3
NW Atlantic
(Runyan, 1961; Dovel, 1963)
Rimicola muscarum
Gobiesocinae
6-8
6-8
14-16
8
35-36
53.2
NE Pacific
(Allen, 1979)
Lepadulilhys frenatus
Diademichthyinae
15-17
12-15
25-31
11
—
52.5
W Pacific
(Shiogaki and Dotsu, 1971b, c)
Aspasma minima
Aspasminae
7-9
6-9
21-24
8-9
—
52.3
NW Pacific
(Shiogaki and Dotsu, 1971a)
Aspasmichthys cicomae
Aspasminae
11-13
8-9
23
10-11
—
56.0
NW Pacific
(Shiogaki and Dotsu, I972d)
gions of the world). The relationships among the subfamilies
were based pnmaiily on four characters: the number of gill
arches; gill membrane state; type of suction disc and dentition
type.
In my opinion, the evolutionary scheme presented in Briggs
(1955) is in drastic need of revision from a cladistic viewpoint.
The independent derivation of the single suction disc and mixed
derived character states in several divergent evolutionary lines,
plus the "evolution" of one subfamily from another through
primitive and derived genera are particularly troubling aspects
of his analysis.
Gosline (1970) was first to include the Callionymidae, Dra-
conettidae and Gobiesocidae in the order Gobiesociformes. Ac-
cording to Gosline (1970) the three families share a number of
characteristics including a scaleless head and body, no circum-
orbital bones behind lacrimal, articular processes of the pre-
maxillae, as well as others (see Gosline, 1970: 365 and 377).
These similarities coupled with evidence that, in Gosline's words,
"the Gobiesocidae has evolved from the notothenoid section of
the perciform suborder Blennioidei and in small part at least
over the same route as the draconettids and callionymids" form
the basis for including all three families in the order Gobiesoc-
iformes.
Greenwood et al, (1966) placed the Gobiesociformes which
included only the Gobiesocidae mto the superorder Paracan-
thopterygii in their provisional classification of teleostean fishes.
Apparently this placement was based on a relationship between
batrachoidids and gobiesocids proposed by Briggs (1955) and
McAllister (1968), although Briggs did note some resemblance
between the Gobiesocidae and the Callionymoidea. Gosline
(1970) believed that characteristics held in common by gobie-
socoid and batrachoid fishes (e.g., the usually scaleless body.
Fig. 336. Representative larvae of seven genera within the Gobiesocidae: (A) Comdens taticephalus. 5.5 mm (from Shiogaki and Dotsu,
1971d); (B) Trachelochismus melobesia. 7.8 mm (after Ruck, 1971); (C) Lepadogaster lepadogaster. 6.0 mm (after Russell, 1976); (D) Apletodon
microcephalus. 4.5 mm (after Russell, 1976); (E) Dtplecogaster bimaculata. 6.5 mm (after Russell, 1976); (F) Diplocrepis puniceus. 7.7 mm (after
Ruck, 1973b); and (G) Gastrocyathus gracilis. 6.9 mm (after Ruck, 1976).
634
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
B
ALLEN: GOBIESOCIFORMES
635
Dorsal Trunk (DT)
Dorsal Head (DH)
Lateral Trunk (LT)
Postanal Ventral (PV)
Dorsal Gut (DG)
Lateral Gut (LG)
Fig. 338. Hypothetical clingfish larva showing regions which form the basis for coding patterns of melanophores.
Ventral Gut (VG)
flattened head, anterior pel vies, incomplete circumorbital series)
are the result of convergence. Gosline (1970) then cited five
morphological and osteological features which differ between
the Gobiesociformes (including the three families) and the Ba-
trachoidiformes. These features include differences in pelvic fin
structure and orientation, structure of the upper hypurals, as-
cending process of premaxilla, ossification of the median eth-
moid and presence (Batrachoidiformes) or absence (Gobiesoc-
iformes) of a swimbladder. On the other hand, he upheld that
gobiesociform (three families) fishes have almost all of the di-
agnostic characteristics of the superfamily Notothenioidca of
the perciform suborder Blennioidea (see Gosline, 1968). He
further pointed out structural similarities between members of
the Gobiesociformes and certain genera of notothenioid fishes
as evidence supporting this proposed relationship. Based on this
work on adults, gobiesociform fishes are currently considered
perciform derivatives in the superorder Acanthopterygii. How-
ever, the issue remains far from resolved and future investi-
gations into both the ordinal and superordinal relationships are
still very much in order. In fact, William Eschmeyer (California
Academy of Sciences) is currently investigating possible rela-
tionships between gobiesociform (particularly gobiesocids) and
scorpaeniform fishes (pers. comm.).
The early life history stages of gobiesocids and callionymids
(see Houde, this volume) lend little support to Gosline's clas-
sification. Gobiesocid and callionymid larvae are usually pig-
mented heavily, but there are very few additional similarities
at the current level of examination. Gobiesocid and callionymid
early life history stages differ in: egg type (demersal versus pe-
lagic eggs, respectively), preanal length (>50% versus <50% of
standard length), general body shape (relatively large cylindrical
versus small, laterally compressed larvae), myomere/vertebral
counts (24 to 37 versus 19 to 23), and shape of the notochord
tip (no extension versus a long extension beyond the hypural
plate). These basic differences may, in part, represent divergence
due to dissimilar reproductive strategies. A more thorough, de-
tailed comparison of the early life history stages (larvae in par-
ticular) will be necessary before any solid conclusions can be
drawn. Unfortunately, the eggs and larvae of draconettids (pre-
sumably the most primitive members of the order) are unknown
and cannot help clarify the situation.
The use of larval characteristics to assess higher level rela-
tionships between the Gobiesociformes and the Batrachoidi-
formes or Notothenioidea is limited since batrachoids have di-
rect development (no larval form) and the larvae of notothenioids
bear little, general resemblance to gobiesocid and callionymid
Fig. 337, Representative larvae of six genera within the Gobiesocidae: (A) Gastroscyphus hectoris. 5.4 mm (after Ruck. 1976); (B) Gobiesox
rhessodnn. 6.2 mm (from Allen, 1979); (C) Rimuola miiscarum. 4.0 mm (from Allen, 1979); (D) Lepadichthys frenalus, 7.3 mm (from Shiogaki
and Dotsu, in prep.); (E) Aspasma minima, 6.8 mm (from Shiogaki and Dotsu, 1971a); and (F) Aspasmichlhys ciconiae. 6.9 mm (from Shiogaki
and Dotsu, 1972d).
636
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Aspasma minima
(double disc)
CO
.8mm
; V 6.9mm
Lepadichthvs frenatus
(single disc)
7.3mm
larvae. Detailed studies with the larval (or embryonic) forms of
the above mentioned groups should be fruitful in leading, I
believe, to a clearer understanding of their relationships.
Future investigation into gobiesociform systematics should
first concentrate on whether the current gobiesociformes rep-
resents a monophyletic grouping. Only when this question is
answered satisfactorily can the higher order relationships be
addressed.
Department of Biology, California State University,
NORTHRIDGE, NORTHRIDGE, CALIFORNIA 91330.
8.3mm
9. 1 mm
8.2mm
¥\ 10. Or
10.0mm
Fig. 339. Development sequences of the pelvic suction disc in larval
gobiesocids (double and single types).
Callionymidae: Development and Relationships
E. D. HouDE
THE Callionymidae are one of three families in the order
Gobiesociformes (Gosline, 1970; Nelson, 1976). System-
atics, ontogeny and relationships of Callionymidae and the other
families, Gobiesocidae and Draconettidae, have been reviewed
and summarized for this symposium by Allen (this volume).
The callionymids are small demersal fishes found in all warm
seas. Most species are less than 100 mm in length. Maximum
length is about 200 mm (Nelson, 1976; Nakabo, 1 982b). Nelson
( 1976) indicated that there are eight genera with about 40 species
in the family. Fncke (1980, 1981a, 1981b) believed the family
to be more diverse with perhaps 130 species worldwide, 75 in
the genus Callionymus. and Nakabo (1982b) recently has pro-
posed 19 genera and 139 species. Callionymids are most abun-
dant and diverse in shallow marine waters of the Indo-Pacific
(Smith, 1963; Fricke, 1980, 1981b). They also are common in
the Atlantic (Davis, 1966). Although usually found in depths
less than 1 00 m, some species occur to depths of > 600 m (Davis,
1 966). Larvae sometimes are abundant in coastal ichthyoplank-
ton surveys. For example, Callionymus pauciradiatus was the
second most common species of larva in Biscayne Bay, Florida
(Houde and Alpem Lovdal, in press) while Callionymidae were
the sixth most abundant family of larvae in Persian Gulf fish
larvae collections.'
Nakabo (1982b) has extensively revised the Callionymidae,
establishing 7 new genera and redefining 12 previously recog-
nized genera/subgenera. Genera are defined based on cephalic
lateral lines, lateral lines on the body, morphology, secondary
' Houde, E. D., J. C. Leak, S. Al-Matar and C. E. Dowd. 1981.
Ichthyoplankton abundance and diversity in the western Arabian Gulf.
Kuwait Institute for Scientific Research. Mariculture and Fishenes De-
partment. Final Report, Project MB- 16, 3 volumes. (This report was
not available for distnbution at the time the present paper was written.)
Table 169.
Meristics
OF Calliony
MID Genera
Recognized
> BY Nakabo (198:
2b). His new gene
ra are designated n.g.
Genus
Dorsal
Anal
Pectoral
Pelvic
Caudal
Type species by onginal designation
Callionymus
Balhycallionymus n.g.
in
-IV,
IV,
6-
9
•10
9
9
i
11
+ 16-20
+ 17-19
I,
I,
5
5
i + 7
i + 3
+ ii
+ ii + 2 + ii
Callionymus lyra
Callionymus kaianus
Foetorepus
Eocallionymus n.g.
Paracallionyinus
Neosyncluropus n.g.
IV,
IV,
IV,
IV,
8
7
9-
8
■10
7
6
9
7
i
i
i
i
+ 18-21
+ 18-19
+ 18-19
+ 17-21
I,
1,
I,
1,
5
5
5
5
i +
i +
i +
i +
1 +
3 + i + 3 + ii
7 -1- ii
7 + ii
7 + 11
7 + 11
Callionymus calauropomus
Callionymus papitio
Callionymus coslatus
Callionymus ocellalus
Pterosynchiropus n.g.
Minysynchiropus n.g.
IV,
IV,
8
9
7
8
ii
iii
ii
+ 17-19
or
+ 18
+ 29-30
+ 16
or
+ 14-15
or
+ 13-14
+ 15-20
I,
I,
5
5
i +
1 +
8
7
+ i
+ ii
Callionymus splendidus
Synchiropus laddi
Paradiplngrammus n.g.
IV.
8-
-9
7-8
iii
iv
i
I,
5
i +
7
+ ii
Callionymus enneaciis
Diplogrammus
IV,
6-
■8
4-7
ii
i
+ 14-15
+ 16-18
or
+ 15-17
+ 18
+ 17-18
+ 17
+ 16-19
+ 17-18
I,
5
i +
7
+ ii
Callionymus goramensis
Synchiropus
Orhonymus
Dactylopus
Calhunchlhys
Pseudocalhunchlhys n.g.
IV,
IV,
IV,
IV,
IV,
8
8
8
9
8
7
7
7
8
7
ii
i
i
ii
II
1
I,
1,
1.
1.
I,
5
5
1-4
5
5
i +
i +
i +
i +
i +
7
7
7
7
7
+ ii
+ ii
+ ii
+ ii
+ ii
Callionymus lateralis
Callionymus rameus
Callionymus daclylopus
Callionymus japonicus
Callionymus variegatus
Repomucenus
Spinicapilichlhys
Anaora
Eleulhcrnchir
III
1
-IV, 9
IV, 8
IV, 8
absent
or
-IV. 9-
■13
8-9
8
7
9-13
ii
i
i
1
+ 14-17
+ 16-21
+ 18-20
21-25
+ 16-23
I,
I.
I,
I.
5
5
5
5
i -t- 7 + ii
i -f 7 + ii
ii + 6 + ii
i + 7 + ii
or li + 6 + ii
Callionymus calcaralus
Callionymus spiniceps
Anaora lenlaculata
Callionymus opercularoides
637
638
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
HOUDE: CALLIONYMIDAE
639
sex characteristics and body size. Meristics also vary among
species (Table 169). In Nakabo's classification the genus Cal-
lionvimis includes only five species, all of which are fiaund in
the northeast Atlantic, Mediterranean or Black seas. Nakabo
(1982b) assigned 39 species to the Indo-Pacific genus Repo-
mucemts. making it the most species-nch genus of Callionym-
idae, followed by the Indo-Pacific genus Calliurichthys with 16
species. Neither eggs nor larvae were described or discussed by
Nakabo m his systematic account of the Callionymidae.
Description
The Callionymidae are characterized by having a small, pore-
like gill opening, greatly reduced in size compared to that of
Draconettidae, their closest relatives in the Gobiesociformes.
The preoperculum has a strong, often serrate, spine, useful for
specific identifications; the operculum and suboperculum are
spineless. Eyes are dorsal and adjacent. Hypurals are fused into
a single plate. Vertebrae number 21 to 23. Dorsal fin spines
usually four; soft rays 6-13; anal fin with 4-13 soft rays. Pelvic
fins are inserted in advance of the pectoral base, the two fins
often connected at their bases by a membrane. The sexes usually
are dimorphic, males having longer and broader median fins,
sometimes with filamentous rays in the dorsal and caudal fins.
Development
Size at maturity varies among species but generally is less
than 100 mm. Some species may mature at < 15 mm in length
(Davis, 1966). Callionymid male-female pairs exhibit well-de-
fined courtship and spawning behavior (Wilson, 1978; Takita
and Okamoto, 1979) m which male display plays a prominent
role. Individual females may spawn on successive days. Judging
from larval occurrences, spawning seasons are protracted, last-
ing 6 months or more for temperate species such as C. lyra. C.
maculatus and C. reticulatus (Demir, 1972; Russell, 1976).
Spawning may occur year-round in subtropical species such as
C. pauciradiatus (Houde and Alpem Lovdal, in press) and Par-
acallionymus coslatus (Brownell, 1979) or tropical species such
as C. decoratus (Watson and Leis, 1974).
Eggs
Eggs are colorless, pelagic and spherical, reported diameters
ranging from 0.55 to 0.97 mm (Mito, 1962a; Watson and Leis,
1974; Russell, 1976; Brownell, 1979; Miller et al., 1979; Takai
and Yoshioka. 1979; Takita, 1980, 1983). A polygonal (usually
hexagonal) sculpturing, sometimes with fine cilia-like processes,
usually is associated with the chorion, but in some species (e.g.,
P. costatus) the chorion apparently is unsculptured (Brownell,
1979). Buoyant, adhesive egg masses have been described for
C. calliste. which break up into individual pelagic eggs prior
to hatching (Takita, 1983). The yolk is segmented peripherally.
The perivitelline space is narrow. There are no oil globules.
Takai and Yoshioka (1979) and Takita (1980) have provided
good illustrations and photographs of typical callionymid eggs.
Larvae
At hatching, pelagic larvae of callionymids range from ap-
proximately 1 .0 to 2. 1 mm in length. Most species are less than
1.5 mm at hatching, making them among the smallest of larval
fishes. Reported myomere numbers range from 19-22. Callio-
nymid larvae are distinctive and easy to recognize. Larvae of
several species (referred to as Callionymus) have been described
(e.g.. Page, 1918; Mito, 1962a; Demir, 1972, 1976; Miller et al.,
1979; Takai and Yoshioka, 1979; Takita, 1980, 1983). Brownell
(1979) has illustrated larvae of Paracallionymus coslatus. All
larvae described to date are similar, differing in pigmentation
patterns, meristic characters and sizes at which fin development
and metamorphosis are completed.
Yolk-sac larvae are short and deep-bodied with a large, bul-
bous yolk sac (Mito, 1962a; Brownell, 1979; Takita, 1980, 1983).
The yolk is segmented peripherally. Dendritic or stellate me-
lanophores may develop in the finfold (Fig. 340B) within one
day after hatching (Mito, 1962a; Brownell, 1979; Takai and
Yoshioka, 1979; Takita, 1980, 1983). The snout-to-anus length
of newly-hatched larvae is >50% of notochord length, but it
declines to <50% within several hours after hatching.
Preflexion larvae are moderately deep-bodied and laterally
compressed both preanally and postanally. All species described
to date have a broken line of melanophores along the lateral
midline, particulariy on the tail (Fig. 340). The larvae are mod-
erately to heavily pigmented and often are first recognized in
samples because of their relatively dark color. A swimbladder
which develops at this stage subsequently is lost during meta-
morphosis. Curious processes, termed "spine-like" by Takita
( 1 980, 1 983) or called "serrations" by Mito ( 1 962a) develop at
the margins of the dorsal and ventral finfolds (Fig. 340A), which
apparently vary in number among individual larvae. Takita
( 1 980) described and illustrated a "vacuole" in the dorsal finfold
of small, preflexion larvae of C. flagris. C. richardsoni and C.
ornalipinnts. Multiple vacuoles were reported in the finfolds of
C. ca/fa/f (Takita, 1983).
Postflexion larvae are heavily pigmented and robust (Fig.
340C). They have a prominent and highly visible, upturned
notochord tip (urostyle). Caudal, pelvic, second dorsal and anal
fin ray counts may be complete in some species at 3-4 mm SL
(Miller et al., 1979; Takai and Yoshioka, 1979). The head be-
comes flatter and broader as development progresses and the
eyes gradually assume their dorsal, adjacent position. The pre-
opercular spine first appears in the length range 3.5 to 5.0 mm
SL. For most species, size at metamorphosis is approximately
10 mm SL.
Relationships
Callionymid eggs and larvae offer little clue to systematic
relationships among gobiesociform fishes. Like the gobiesocids,
callionymid larvae are heavily pigmented (Allen, this volume)
but there are few additional similarities. Callionymid larvae
hatch from pelagic eggs; gobiesocids have demersal eggs. From
Fig. 340. Larvae of Callionymidae: (A) 1.7 mm larva of Callionymus (Paradiplogrammus) calliste (from Takita, 1983: fig. 21, p. 443); (B) 4.7
mm larva of Callwnvmus reticulatus (from Demir. 1972: fig. 2. p. 998); (C) 4.1 mm lar\'a of Callionymus (Repomucenus) beniteguri {from Takai
and Yoshioka, 1979: fig. 2-4, p. 150); (D) 2.9 mm larva of Callionymus (Calliurichthys) decoratus (from Miller et al., 1979: fig. 96, p. 96); and
(E) 2.3 mm larva of Paracallionymus costatus (from Brownell, 1979: fig. 69, p. 50).
640
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
several hours after hatching and during preflexion stages, most
species of caliionymid larvae have snout-to-vent lengths <50%
of standard length, while gobiesocids have snout-to- vent lengths
>50%. Both callionymids and gobiesocid larvae have swim-
bladders which are lost during metamorphosis, a characteristic
common to many teleost families. Caliionymid preflexion larvae
are small and laterally compressed. They have low myomere/
vertebral numbers (19 to 23). Gobiesocid larvae are relatively
large, basically cylindrical in shape, with high myomere/verte-
bral counts (24 to 37) (Allen, this volume). The various species
of caliionymid larvae differ little from each other but they can
be identified by distinctive pigment patterns and median fin ray
counts (Demir, 1972; Miller et al., 1979: Takai and Yoshioka,
1979; Takita, 1980, 1983; Olney and Sedberry, in press). There
has been no attempt yet to relate larval characters or differences
among larvae to the generic characters proposed by Nakabo
(1982b). There are no descriptions of eggs or larvae of Draco-
nettidae, adults of which bear close resemblance to Calliony-
midae (Davis, 1966). The discovery and description of larval
draconettids might resolve the systematic uncertainties among
gobiesociform fishes. A careful, comparative analysis of cal-
iionymid larval development may clarify the generic relation-
ships among species within Callionymidae.
University of Maryland, Center for Environmental and
EsTUARiNE Studies, Chesapeake Biological Laboratory,
Solomons, Maryland 20688.
Pleuronectiformes: Development
E. H. Ahlstrom, K. Amaoka, D. A. Hensley,
H. G. MOSER AND B. Y. SUMIDA
PLEURONECTIFORM fishes have both eyes on one side of
the head in juveniles and adults. The eyes are symmetrical
in larvae, and migration of either the left or right eye occurs
during metamorphosis. In some flatfish groups the eyes are on
the left side (sinistral) while in others they are on the right side
(dextral); relatively few species are indiscriminate. In some flat-
fishes the ocular nerve of the migrating eye usually lies dorsal
to the other nerve in the optic chiasma; in other groups the
nerve of the migrating eye is dorsal or ventral in the chiasma
with about equal frequency. In most groups the nasal organ of
the blind side also migrates to a position near the dorsal midline.
Features of the dentition and cranial osteology may also show
asymmetry. Flatfishes are highly compressed with the underside
of the body usually unpigmented. The lateral line may be lacking
on the blind side; the pectoral fin is often shorter on the blind
side and has fewer rays; the pelvic fin on the blind side is often
shorter, smaller and diflierently placed with respect to the ventral
midline compared with the pelvic fin on the ocular side; squa-
mation may be different on the two sides of the body. The dorsal
and anal fins are long-based; the dorsal extends anteriad to at
least the eye in all flatfishes except Psettodes and the anal fin
extends well forward of the first haemal spine. The caudal fin
is typically rounded or truncate with few or no secondary rays.
Pleuronectiforms are benthic carnivores, occurring worldwide,
primarily in shallow to moderate depths, with some represen-
tatives in brackish and fresh water habitats. Nelson (1976) notes
a total of 520 species.
The classification presented below is based on the works of
Regan (1910, 1929) and Norman (1934, 1966) with modifica-
tions by Hubbs (1945), Amaoka (1969), Hensley (1977), and
Futch ( 1 977). Our removal of Perissias from the Paralichthyidae
and placement in the Bothidae are based on previously unpub-
lished information. Those genera marked with an asterisk are
misplaced in this classification and are discussed in this paper
and in Hensley and Ahlstrom (this volume).
Order Pleuronectiformes
Suborder Psettodoidei
Family Psettodidae (Indo-Pacific, West Africa)
Pseltodes
Suborder Pleuronectoidei
Family Citharidae
Subfamily Brachypleurinae (Indo-Pacific)
Brachyplcura* Lepidohlepharon
Subfamily Citharinae (Indo-Pacific, Mediterranean,
West Africa)
Citharoides, Euatharus
Family Scophthalmidae (North Atlantic, Mediterranean,
Black Sea)
Lepidorhombus. Phrynorhombus, Scophthalmus,
Zeugoplerus
Family Paralichthyidae (Western and Eastern Atlantic,
Eastern Pacific, Indo-Pacific)
Ancylopsetta, Cephalopsctta, Cithanchthys, Cyclop-
Fig. 341. Eggs of Pleuronectiformes. Captions in each illustration indicate the species and diameter of the egg in mm. Scophthalmus maeoticus
maeoticus. from Dekhnik, 1973; Paralichthys oltvaceus. from Mito, 1963; Bothidae, from Mito, 1963; Limanda aspera. from Pertseva-Ostroumova,
1954; Hippoglossotdes duhius, from Pertseva-Ostroumova, 1961; Microstomus pacijicus. onginal, CalCOFI; Pleuronichlhys cornutus. from Mito.
1963; PehlretisJIavilalus. from Robertson. 1975a; Pellorhamphus novaezeelandiae. from Robertson. 1975a; Tnnecles maculalus. from Hildebrand
and Cable, 1938; Pegusa lascans nasula. from Dekhnik, 1973; Cynoglossus robuslus. from Fujita and Uchida, 1957.
AHLSTROM ET AL.: PLEURONECTIFORMES
641
1.10- 1.33
Scophthalmus maeotlcus
maeoticus
0.92
Paralichthys olivaceus
0.64
Bothidae
0.76 - 0.85
2.10 - 2.94
2.05 - 2.57
Limanda aspera
Hippoglossoides dubius
Microstomus pacificus
1.22
0.62 - 0.68
0.62 - 0.68
Pleuronichthys cornutus
Pelotretis flavilatus
Peltorhamphus
novaezeelandiae
0.67 - 0.86
1.09- 1.35
0.85 - 0.90
Trinectes maculatus
Pegusa lascaris nasuta
Cynoglossus robustus
642
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 70. Characters of Eggs of Pleuronectinae Species which Lack Oil Globules.
Type of egg
(pelagic or
Egg size
Species
Region'
demersal)
(mm)
Chonon
References
Cteisthenes herzensteini
WNP
P
0.84-1.03
Pertseva-Ostroumova, 1 96 1
Embassichlhys balhybius
ENP
P
3.0
smooth
Richardson, 1981b
Eopsella grigorjewi
WNP
P
1.10-1.20
striations
Yusa, 1961; Fujita, 1965
E. jordani
ENP
P
1.21-1.25
smooth
Alderdice and Forrester, 1971
Glyptocephalus cynoglossus
NA
P
1.07-1.25
striations
Cunningham, 1887; Ehrenbaum,
1905-1909
G. slelleri
WNP
P
1.20-1.61
thick, reticulate
Pertseva-Ostroumova, 1961;
Dekhnik, 1959
G. zachirus
ENP
P
1.9-2.15
striations
Original
Hippogtossoides dubius
WNP
P
2.10-2.94
smooth
Pertseva-Ostroumova, 1961
H. elassodon
NP
P
2.45-3.75
smooth
Thompson and Van Cleve, 1936;
Pertseva-Ostroumova, 1961
H. platessoides
WNA
P
1.38-2.64
smooth
Russell, 1976 (summary)
H. robuslus
WNP
P
2.04-2.69
smooth, thin
Periseva-Ostroumova, 1961
Hippoglossus hippoglossus
NA
P
3.0-3.8
smooth, thick
Tuning, 1936; Pertseva-Ostroumova,
1961
Thompson and Van Cleve, 1936
H. stenolepis
NP
P
2.9-3.8
minute honey-
comb structure
Isopsetta isolepis
ENP
P
0.90-1.10
smooth
Richardson et al., 1980
Kareius bicoloratus
WNP
P
1.00-1.15
reticulate
Pertseva-Ostroumova, 1961
Lepidopsetta bilineala
NP
D
1.02-1.09
sticky-orange
Pertseva-Ostroumova, 1961
L. inochigarei
WNP
D
0.87-0.95
thick, gluey
Pertseva-Ostroumova, 1961
Limanda aspera
NP
P
0.76-0.85
smooth
Pertseva-Ostroumova, 1954
L. ferruginea
WNA
P
0.79-1.01
striations
Miller, 1958; Colton and Marak,
1969
L. limanda
ENA
P
0.66-1.20
—
Russell, 1976 (summary)
L. punctatissima
WNP
P
0.66-0.87
smooth
Pertseva-Ostroumova, 1961
L. proboscidea
WNP
P
0.72-0.87
smooth
Pertseva-Ostroumova, 1961
L. schrenki
WNP
D
0.73-0.83
adhesive
Yusa, 1960a
L. schrenki (as Pseudopleuronectes
WNP
D
0.74-0.83
adhesive
Pertseva-Ostroumova, 1961
yokohamae)
L. yokohamae
WNP
D
0.81-0.84
adhesive
Yusa, 1960a, b
Liopsetta glacialis
NP
P
1.20-1.60
thin
Pertseva-Ostroumova, 1961
L. obsciira
WNP
D
0.78-0.94
thick, sticky
Pertseva-Ostroumova, 1961
L. pinnifasaata
WNP
P
1.43-1.66
thin, folds
Pertseva-Ostroumova, 1961
Lyopsella exihs
ENP
P
1.47-1.68
smooth
Original; Ahlstrom and Moser, 1975
Microslomus kilt
WNA
P
1.13-1.45
striations
Russell, 1976; Dekhnik, 1959
M. pacificus
ENP
P
2.05-2.57
smooth
Original; Ahlstrom and Moser, 1975
Parophrys vetulus
ENP
P
0.89-0.93
striations
Budd, 1940; Original
Platichthys flesus
ENA
P
0.80-1.13
_
Russell, 1976 (summary)
P. f. luscus
B
P
1.05-1.35
smooth
Dekhnik, 1973
p. slellatus
NP
P
0.89-1.01
smooth, thin
Orcutt, 1950; Yusa, 1957
Pleuronecles pallasii
NP
P
1.67-2.21
—
Pertseva-Ostroumova, 1961
P. ptatessa
ENA
P
1.66-2.17
—
Russell, 1976 (summary)
Pleuronichlhys coenosus
ENP
P
1.20-1.56
polygonal
pattern
Sumidaetal., 1979; Budd, 1940
P. decunens
ENP
P
1.84-2.08
polygonal
pattern
Sumida et al., 1979; Budd. 1940
P. verlicalis
ENP
P
1.00-1.16
polygonal
pattern
Sumida et al., 1979; Budd, 1940
Psellichlhys melanosticlus
ENP
P
ca. 1.0
—
Hickman, 1959
Pseudopleuronectes amertcanus
WNA
D
0.71-0.96
adhesive
Breder, 1923
P. herzensteini
WNP
P
0.80-1.0
smooth
Pertseva-Ostroumova, 1961
Reinhardlius hippoglossoides
NA/NP
P
4.00-4.50
—
Jensen. 1935
Tanakius kitaharai
WNP
P
1.20-1.30
striations
Fujita, 1965
Verasper vanegatus
WNP
P
1.60-1.64
smooth
Takitaetal., 1967; Mito, 1963
' B = Black Sea, ENA
Pacific.
= eastern North Atlantic. ENP = eastern North Pacific, I = India, NA = North Atlantic. NP = North Pacific. WNA = western North Atlantic. WNP = western North
setta. Etropus. Gastropsetta. Hippoglossina.
Lioglossina. Paralichthys. Pseudorhombus,
Syaciurn. Tarphops, Tephrinectes* Thysanop-
setta* Verecundum. Xystreurys
Family Bothidae
Subfamily Taeniopsettinae (Western Atlantic, Eastern
Pacific, Indo-Pacific)
Engyophrys. Perissias, Taeniopsetta, Trichopsetta
Subfamily Bothinae (Indian, Pacific, Atlantic, Medi-
terranean, SoutheiTt oceans)
Arnoglossus. Asterorhoinhiis, Bothus. Chascanopset-
ta, Crossorhoinbus. Engyprosopon. Grainma-
lobothus, Japonolaeops. Kamoharaia, Laeops,
Lophonectes, Monolene, Mancopsetta* Neo-
AHLSTROM ET AL.: PLEURONECTIFORMES
643
Table 171. Characters of Pleuronectiform Eggs with a Single Oil Globule.
Egg size
Oil globule
Size at
Taxon
Region'
(mm)
si/e (mm)
Chorion
halching (mm)
References
Pleuronectidae
Hypsopsella gutlulala
ENP
0.78-0.89
0.12-0.14
smooth
1.7-2.3
Sumidaet al., 1979;
Eldridge, 1975
PIcuromchthys cornutus
WNP
1.03-1.25
polygonal
pattern
2.8-3.8
Mito, 1963; Takita and
Fujita, 1964
P. niieri
ENP
0.94-1.08
0.08-0.14
polygonal
pattern
2.1
Sumidaet al., 1979
Scophthalmidae
Lepidorhombus whiffiagonis
ENA
1.02-1.22
0.25-0.30
striations
ca. 4.0
Mcintosh, 1892; Holt,
1893
Phrynorhombus norvegicus
ENA
0.72-0.92
0.09-0.16
rugose
2.5-2.8
Ehrenbaum, 1905-1909;
Hefford, 1910
P. regius
ENA
0.90-0.99
0.16-0.18
_
2.4
Holt, 1897
Scophthalmus aquosus
WNA
0.90-1.38
0.15-0.30
striations
ca. 2.0
Martin and Drewry, 1978
(summary)
S. maeoticus
B
1.10-1.33
0.17-0.23
_
3.5 TL
Dekhnik, 1973
S. mcLximus
ENA
0.90-1.20
0.15-0.22
rugose
2.1-3.0
Holt, 1892; Jones. 1972
S. rhombus
B
1.20-1.50
0.16-0.25
striations
3.8
Jones, 1972
Zeugoplerus punclatus
ENA
0.92-1.07
0.17-0.20
-
2.5-2.9
Hefford, 1910
Paralichthyidae
llippoglossina obtonga
WNA
0.91-1.12
0.17
smooth
2.7-3.2
Miller and Marak, 1962
H. slomala
ENP
1.22-1.38
0.20-0.26
smooth
3.7
Sumida et al., 1979
Paralichlhys californicus
ENP
0.74-0.82
0.10-0.19
smooth
ca. 2.0
Original
P. dentatus
WNA
0.90-1.10
0.18-0.31
_
2.4-2.8
Smith and Fahay, 1970
P. olivaceus
WNP
0.83-1.03
0.13-0.21
smooth
2.6-2.8
Mito, 1963
Pseudorhombus cinnamoneus
WNP
0.77-0.89
0.12-0.14
_
1.8-2.0
Mito, 1963
Cilhanchlhys arclifrons
WNA
0.70-0.82
—
smooth
ca. 2.0
Richardson and Joseph,
1973
Bothidae
Arnoglossus capensis
ESA
0.72
0.12
smooth
2.2
Brownell, 1979
A. kessleri
B
0.59-0.70
0.10-0.13
smooth
1.8-1.9
Dekhnik. 1973
A. taterna
ENA
0.60-0.76
0.11-0.15
smooth
2.6
Russell, 1976 (summary)
A. scapha
NZ
0.78-0.88
0.11-0.12
smooth
—
Robertson, 1975a
A. ihon
ENA
0.67-0.74
0.12
smooth
1.6-2.0
Russell, 1976 (summary)
' ESA = eastem South Atlantic; NZ = New Zealand; key to other regions as in Table 1 70.
laeops. Parabothus. Pelecanichthys. Psettina,
Tosarhoinbus
Family Pleuronectidae
Subfamily Pleuronectinae (Atlantic, Mediterranean,
Pacific, Arctic)
Acanthopsetta, Atheresthes. Cleisthenes. Clidoder-
ma. Dexistes, Embassichthys, Eopsetta. Glyp-
tocephalus, Hippoglossoides. Hippoglossus,
Hypsopsella. Isopsetta. Lepidopsetta. Limanda.
Liopseila. Lyopsetla, Mtcrostomus. Parophrys.
Platichlhys. Pleuronectes, Pleuromchthys. Psel-
tichthys. Pseudopleuronectes. Reinhardlius.
Tanakius. Verasper
Subfamily Poecilopsettinae (Indo-Pacific, Atlantic)
Marleyclla. Nematops. Poecilopselta
Subfamily Paralichthodinae (Indian Ocean off South
Africa)
Paralichthodes
Subfamily Samarinae (Indo-Pacific)
Samans, Samariscus
Subfamily Rhombosoleinae (New Zealand, Southern
Australia, South America)
Ammolretis. Azygopus. Colislium, Oncopterus. Pe-
lotretis. Peltorhamphus. Psammodisciis. Rhom-
bosolea
Suborder Soleoidei
Family Soleidae
Subfamily Soleinae (Worldwide, tropical to temperate)
Norman (1966) recognized 22 genera
Subfamily Achirinae (American coasts, some fresh
water) Norman (1966) recognized 9 genera
Family Cynoglossidae
Subfamily Symphurinae (Tropical-Subtropical Amer-
ican coasts, Mediterranean, West Africa, Indo-Pa-
cific)
Symphurus
Subfamily Cynoglossinae (Indo-Pacific, Mediterra-
nean, West Africa, Japan, some fresh water)
Cynoglossus, Paraplagusia
A profuse literature on the life history stages of flatfishes has
accumulated since the early work of Cunningham (1887, 1889,
1890, 1891) who described numerous series reared from eggs
collected from running ripe females. Other European workers
(Holt. 1893; Mcintosh and Pnnce, 1890; Petersen, 1904, 1905,
1906, 1909; Schmidt, 1904; Kyle, 1913) identified early life
history series of additional species so that, by the time of Eh-
renbaum's (1905-1909) summary, ontogenetic stages of a major
portion of the eastem North Atlantic flatfish fauna were known.
Padoa (1956k) summarized ontogenetic information on Medi-
644
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 172. Characters of Pleuronectiform Eggs with Multiple Oil Globules.
Egg size
Region'
(mm)
Number of oil globules
Yolk
Chonon
References
Achirinae
Achirus lineatus
WNA
0.71-0.76
12-14
homogeneous
smooth, thin
Houdeet al., 1970
Trinecles maculaius
WNA
0.67-0.86
15-34
homogeneous
smooth
Hildebrand and Cable.
1938
Soleinae
Aesopia cornula
WNP
1.45-1.60
many; scattered
homogeneous
polygonal mesh
Mito. 1963
Auslroglossus microlepis
ESA
0.88
12-20
homogeneous
smooth
Brownell, 1979
Buglossidiuni luieum
ENA
0.64-0.94
12-21; scattered
peripheral
smooth
Holt, 1891; Hefford,
1910
Dicologoglossa cuneata
ENA
0.70-0.84
60-80; scattered
peripheral
segmentation
Ugardere, 1980
Microchirus ocellatus
ENA
0.90-1.10
30-40; scattered
penpheral
segmentation
smooth
Palomera and Rubies,
1977
M. variegatus
ENA
1.28-1.42
to 50 + ; scattered
peripheral
segmentation
smooth
Cunningham, 1889
Pegiisa impar
M
1.06
Padoa, 1956k
P. lascaris lascaris
ENA
1.28-1.38
to 50 +
peripheral
segmentation
Holt, 1891; Hefford,
1910
P. lascaris nasuta
B
1.36-1.38
many; clumped
homogeneous
polygonal mesh
Dekhnik. 1973
Solea solea
ENA
0.98-1.58
many; highly
clumped
peripheral
segmentation
smooth
Cunningham, 1889;
Fabre-Domergue and
Bietnx, 1905
Synaptura kleiiu
ESA
1.34
many; clumped
homogeneous
smooth
Brownell, 1979
Zebrias japomcus
WNP
1.75
many; scattered
homogeneous
smooth
Mito, 1963
Z. zehra
WNP
1.60
many; scattered
homogeneous
Mito, 1963
Symphunnae
Symphurus atricauda
ENP
0.71-0.78
10-23
homogeneous
smooth, colored
Onginal
Cynoglossinae
Cynoglossus capensis
ESA
0.75
2-16
homogeneous
smooth
Brownell, 1979
C. robustus
WNP
0.85-0.90
5-15
homogeneous
fine hexagonal
network
Fujita and Uchida, 1957
C. {Areliscus) tngrammus
WNP
1.19-1.23
30-50
homogeneous
smooth
Fujita and Takita, 1965
Cynoglossidae no. 5
WNP
0.71
14
homogeneous
smooth
Mito, 1963
Cynoglossidae sp. A
I
0.84
13-15
homogeneous
smooth
Vijayaraghavan, 1957
Cynoglossidae sp. B
I
0.82
1 8-22; clustered
homogeneous
smooth
Vijayaraghavan, 1957
Cynoglossus I
I
0.60
16-30
homogeneous
smooth
Nair, 1952a
Cynoglossidae (as Solea
I
0.61-0.71
17-25
homogeneous
smooth
John, 1951b
ovaia)
Rhombosoleinae
Ammotrelis rostratus
NZ
ca. 0.8
8-11
homogeneous
smooth
Thomson, 1906
Colistium gunlheri
NZ
1.0-1.08
14-26
homogeneous
Robertson. 1975a
C. nudipinnis
NZ
ca. 1.5
21-28
homogeneous
Robertson, 1975a
Pelotretis JIavilaliis
NZ
0.85-0.95
8-18
homogeneous
smooth
Robertson, 1975a
Peltorhamphus novaczee-
NZ
0.62-0.68
2-6
homogeneous
smooth
Robertson, 1975a
landiae
P. tenuis
NZ
0.58-0.68
2-4
homogeneous
smooth
Robertson. 1975a
Rhombosolea Icporina
NZ
0.58-0.70
2-7
homogeneous
smooth
Robertson, 1975a
R. pleheia
NZ
0.58-0.72
2-13
homogeneous
smooth
Robertson, 1975a; Rob-
ertson and Raj, 1971
Bothinae
Mancopsetta maculala
S
2.45-3.00
20 +
homogeneous
smooth
Efremenko et al., 1981
anlarclica
' M = Mediterranean. S = southern oceans, key to other regions as in Table I 70,
terranean flatfishes and more recently Russell (1976) provided
an extensive review of previous European contributions.
Knowledge of ontogenetic stages of western Atlantic flatfishes
is summarized by Martin and Drewry ( 1 978) and Fahay ( 1 983).
Early life histories of North Pacific flatfishes are treated com-
prehensively by Pertseva-Ostroumova { 1 96 1 ). Japanese and In-
dian workers have provided a long list of contnbutions to flatfish
life history studies and Amaoka (1969, 1979), Hensley (1977)
and Futch (1977) employed ontogenetic characters in assessing
phylogenetic relationships. The individual contributions to flat-
fish ontogeny are too numerous to summarize concisely and are
cited in the section that follows.
Development
Eggs
Eggs are known for most species in Pleuronectidae and
Scophthalmidae and for only a few to moderate numbers of
Fig. 342, Larvae of Psettodidae and Citharidae. (A) Psellodes erumei, 4.3 mm. from Leis and Rennis, 1983: (B) P. eruinei. 8.7 mm. ibid; (C)
Brachypleura novaezeelandiae. 5.0 mm, from Pertseva-Ostroumova, 1965; (D) B. novaezeetandiae. 7.5 mm, ibid.
646
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
.«!»-*■
M. :^r..M
^^^
B
f
D /
-^
V
\\N ^
\\
V---'
'-•^6<:<;^
^
,i-.- s*
A />- '"^ -^
^;;-'5*^«
^ * '" i-fi':' -."♦J
^vl*-'
t;.
Fig. 343. Larvae of Scophthalmidae. (A) Zeugopteriis punctatus. 8.9 mm, dorsal view, from Petersen, 1909; (B) Z. punclalus. 9.0 mm, ibid;
(C) Lepidorhombus boscii. 9.7 mm, ibid; (D) Phrynorhombits regius. 8.0 mm, ibid; (E) Scophlhalmus maximus. 7.4 mm, from Jones, 1972; (F)
5. rhombus. 8.0 mm, ibid.
Fig. 344. Larvae of Paralichthyidae. (A) Paralichthys californkus. 7.0 mm. original, CalCOFI; (B) As above, dorsal view; (C) Xvstreurvs
liolepis. 6.7 mm, original, CalCOFI; (D) As above, dorsal view; (E) Hippoglossina stomata. 8.6 mm, from Sumida et al., 1 979; (F) Pseudorhombus
penlophlhalmus, 9.2 mm, from Okiyama, 1974a; (G) Tarphops oligolepis. 9.2 mm, ibid.
AHLSTROM ET AL.: PLEURONECTIFORMES
647
648
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 173. Summary of Ontogenetic Characters of Pleuronectiforms. (Line indicates data unavailable or presented elsewhere in column.)
Eggs
Lar\'ae
Oil globule:
Chonon:
Single/
Yolk:
Smooth/
Size at
Elongate dorsal
Egg size
Multiple/
Homogeneous/
Slnated/
hatehing
Size at meta-
rays: Present
Taxon
(mm)
Absent
Segmented
Polygonal
(mm)
morphosis (mm)
(no.)/ Absent
Psettodidae
—
—
—
—
—
9-10
10
Citharidae
_
—
_
_
_
_
_
Brachypleura
-
-
-
-
-
7-8
ca. 6
Scophthalmidae
0.72-1.5
s
H
St
2.0-4.0
8->20
A
Paralichthyidae
<0.70-1.38
s
H
s
1.8-3.7
—
—
Paralichthys group
—
—
—
—
—
7.5-15
4-5
Pseudorhomhus group
—
—
—
-
—
8.7-12.5
5-9
Cyclopsetta group
-
-
-
-
-
9->35
*«
Bothidae
0.59-0.88
s
H
s
1.8-2.6
_
_
Taeniopsettinae
—
—
—
—
—
19-60
1
Bothinae
-
-
-
-
-
15-120
1
Pleuronectidae
Pleuronectinae
0.66-4.5
A*
H
S, P. St
1.7-16.0
4.4-65
A
Poecilopsettinae
—
—
—
—
—
ca. 30
A
Samarinae
—
—
—
—
—
ca. 30
1
Rhombosoleinae
0.58-1.5
M
H
s
ca. 1.8
-
-
Soleidae
0.64-1.75
M
H,S
S. P
1.6-4.1
_
_
Soleinae
_
—
—
—
—
3.4-18
A
Achirinae
-
-
-
-
-
3-5.5
***
Cynoglossidae
0.60-1.23
M
H
S, P
1.3-3.2
—
—
Cynoglossinae
—
—
—
—
—
ca. 4-18
2
Symphurinae
—
—
—
—
—
ca. 12-32
usually
4-5
• Single oil globule present in 3 speeies-
** 0-2 in EtropiLi, 0. 2. 3 in Cnhanchthys; 5-8 in Syactum: 8-1 1 in Cyclopsena.
••* Third ray elongated in Achirus.
■••• 0. I. or 2 in Cithanchthy^. 2 or 3 in Cyclopsetta and Syacium: 0 or I in Etropus.
•*••* S Iti^tlata develops elongate third and fourth rays,
****** Protruding in Chascanupsetta. Petecanichthys. and Kamuharaia.
species in other groups, including Soleidae, Cynoglossidae,
Paralichthyidae, and Bothidae.
With a few exceptions, the eggs of flatfishes are pelagic, round,
have homogeneous yolk, a narrow perivitelline space, and an
unsculptured chorion (Fig. 341). The eggs of all flatfishes are
spawned separately. The characters of eggs showing greatest
differences among flatfishes are 1) egg size, and 2) the presence
or absence of an oil globule(s) (Tables 170-172).
Of the approximately 60 species of pleuronectine flatfishes of
the North Pacific and North Atlantic, eggs are known for at least
45 (Table 170). Six species are known to have demersal eggs;
these are round or occasionally off-round and have a sticky,
adhesive chorion that permits clustering or adhesion to bottom
objects. Egg diameters range from 0.66 to 4.5 mm within the
subfamily. The yolk is homogeneous in all pleuronectine eggs.
The perivitelhne space is narrow to moderate, except for eggs
of Hippoglossoides, which have a wide perivitelline space, usu-
ally 25-30% of the egg diameter on either side of the yolk mass.
The chorion has the appearance of being smooth on eggs of most
species, but closer inspection reveals striations or reticulations
on the chorion of some kinds. The chorion of Pleuromchthys
eggs has a striking hexagonal pattern. The eggs of pleuronectine
flatfishes, except for three species, lack an oil globule. The state
of embryonic development achieved in the egg is related to egg
size, more specifically to yolk size. Larvae hatching from small
eggs lack eye pigment, a functional mouth and pectoral fins;
those hatching from larger eggs are much more advanced, with
pigmented eyes, a functional mouth and pectorals. Embryos in
middle- and late-stage eggs are pigmented, with patterns varying
between genera and species. Among species, yolk pigment can
range from unpigmented, to some pigment on yolk adjacent to
the embryo, to heavily pigmented. Pigment can also be present
on finfolds of late-stage eggs of some flatfishes.
Eggs of Scophthalmidae. Paralichthyidae and Bothidae have
a single, small to moderate-sized oil globule, are pelagic, round,
have a narrow to moderate perivitelline space, and homoge-
neous yolk (Fig. 341, Table 171). In late-stage eggs and newly
hatched larvae the single oil globule usually is in the rear of the
yolk mass.
Eggs are known for 8 of the 10 species of scophthalmid flat-
fishes. They range in size from 0.72 to 1.50 mm. The chorion
is striated or rugose in six species and this may apply to all.
Embryos develop considerable pigment over the head and body
and often in finfolds; pigment over the yolk mass and oil globule
can range from none, or sparse, to intense.
Eggs are known for only a few species in the family Paralich-
thyidae. These range in size from 0.70-1.38 mm; chorions are
unsculptured. Except for a few species of Arnoglossus. eggs of
bothid flatfishes are practically unknown. Mito (1963) lists 10
kinds of bothid eggs oflT Japan, unidentified to genus; 8 of these
have diameters under 1.0 mm. Eggs of his Bothidae No. 9 are
slightly ofl"-round and three different eggs have a conspicuous
wart-like appendage. Much work remains to be done in iden-
tifying eggs of fishes of these families, preferably through rearing
eggs from known parents.
Eggs with multiple oil globules are typical of the families
AHLSTROM ET AL.: PLEURONECTIFORMES
649
Table 173. Extended.
Larvae
Elongate
pelvic rays:
Present Absent
Gul:
Normal/
Protruding/
Trailing
Preopercular
spines: Present
Absent
Otic region
spines; Present/
Absent
Frontal region
spines: Present/
Absent
Urohyal spines:
Present/ Absent
Basipter>'gial
spines: Present/
Absent
Cleithral
spines: PrescnL
Absent
Body spines:
Present/Absent
A
N
P
A
A
A
A
A
A
_
P
N
p
A
A
A
A
A
A
A
N
P, A
P. A
P,A
A
A
A
A
—
_
_
_
—
A
N
P
A,P
A
A
A
A
A
A
N
P
A,P
A
A
A
A
A
****
N
P
P,A
A
A
A
A
A
—
—
A
N
A
P
A
P
P
P
A
A
******
A
A
A
P,A
P,A
P,A
P, A
A
N
A,P
A,P
A, P
A
A
A
A
A
N
A
A
A
A
A
A
A
A
P
A
A
A
A
A
A
A
A
N, P
A
A
A
A
A
A
A
A
N. P
A. P
A, P
A,P
A
A
A
P. A
A
P
A
A
A
A
A
A
A
A, *""
P. T
A
A
A
A
A
A
A
Soleidae. Cynoglossidae. the pleuronectid subfamily Rhombo-
soleinae, and Mancopsetia. previously considered a bothid (Fig.
341. Table 172). Eggs have been described for about a dozen
kinds of soleids ranging in size from 0.64-1.75 mm. Eggs are
round, or occasionally slightly off-round. Oil globules are usually
numerous but vary in number, size and distribution within the
yolk. They can be highly clumped, as in Solea solea. or scattered
throughout the yolk, as in Microchirus variegalus. In eggs of the
latter, oil globules were observed to range in size from 0.015-
0.12 mm. whereas they are much smaller and more uniform in
size in Solea solea or Pegusa lascans. Eggs of the two achirine
soleids described from the western Atlantic have a relatively
low number of oil globules. Perivitelline space is narrow to
negligible in soleid eggs. The yolk is peripherally segmented in
eggs of the four species known from the eastern North Atlantic.
Yolk is more completely segmented in the egg designated as
Synapturinae No. 1 by Mito (1963). Yolk can remain unseg-
mented, however, as for example in Achirus lineatus and Tri-
nectes maculatus. Although the chorion of soleid eggs is usually
smooth and unsculptured, Mito (1963) found eggs of Aesopia
cormita to have a pattern of large hexagonal meshes. 0.18-0.24
mm wide, covering the chorion, and Dekhnik (1973) shows fine
polygonal sculpturing on the chorion of P. lascans.
Eggs of the few cynoglossid species known (Table 172) are
small, have homogeneous yolk without secondary segmentation,
a narrow perivitelline space and either an unsculptured chonon
or one with small polygonal meshes. Oil globules range in num-
ber between 5-50, and can be variously distributed in the yolk.
Robertson (1975a) descnbed eggs of seven species of Rhom-
bosoleinae, belonging to four genera (Table 1 72). Egg diameters
range from 0.58 to 1.5 mm. Oil globules in described eggs range
in number from 2-28. Yolk is homogeneous, the perivitelline
space IS narrow, and the chonon is smooth.
Efremenko et al. ( 1 98 1 ) described the ovarian and planktonic
eggs of Mancopsetia macidala antarctica and showed that they
are large (2.45-2.75 mm) and have multiple oil globules (>20).
This finding provides evidence that Mancopsetta does not be-
long in the Bothidae.
Larvae
In addition to such features as meristics, fin arrangement, and
osteology of the fin supports and axial skeleton (which develop
gradually during ontogeny and are essential for identification of
flatfish larvae) the larval stage itself provides many characters
useful in identification and systematic analysis. Larval charac-
ters are summarized in Table 173 and below.
Psettodidae (Fig. 342). — Aboussouan's (1972c) description of
prcflexion larvae of Psettodes bennetti was based on five spec-
imens. 4.4-5.7 mm in length. Leis and Rennis (1983) describe
a series of five larval specimens of Psettodes erumei. 3.0-8.7
mm in length. The smallest specimen has a large yolk sac. the
6.0-mm larva is in mid-flexion and the largest specimen is
undergoing eye migration. Larvae have: a deep, relatively thick
body; large head with massive jaws that extend well beyond the
rear margin of the eye and bear large, early-forming cursed teeth;
large eye; small preopercular spines; and 1 0 early-forming elon-
gate dorsal rays. Dorsal and anal fin rays are all present at 6.0
mm but rays do not appear in paired fins until about 8.0 mm.
Prcflexion larvae have a series of large melanophores along the
dorsal midline, large melanophores alternating with smaller ones
650
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
.jf(im£Mm!^y^^-
Fig. 345. Larvae of Paralichthyidae. (A) Citharichthys stigmaeus. 14.8 mm, from Ahlstrom and Moser, 1975; (B) C. sordtdus, 14.5 mm, ibid;
(C) C. plalophrys. 8.6 mm, onginal, CalCOFI; (D) Etropus crossotus. 6.0 mm, from Tucker. 1982; (E) Cyclopselta chitlendem. 13.0 mm, from
Evseenko, 1982a; (F) Syacium ovale. 6.5 mm, original, CalCOFI.
AHLSTROM ET AL.: PLEURONECTIFORMES
651
Table 174. Numbers of Rays in Dorsal Crest and Size at Developmental Events in Paralichthyidae.
Number of
elongate
dorsal fin
Si/c al
Si/e al
Size at
Species
rays
hatching (mm)
flexion (mm)
Iransfoimation (mm)
References
Hippoglossina stomata
6
3.7
6.2-8.8
9.1->11.7
Sumitia et al., 1979
H ohionga
~6
2.7-3.2
6.3-7.7
10-14
Leonard. 1971
Paralichlhys califormcus
5
2.0
6.0-7.3
7.5-9.4
Original, Ahlslrom and Moser,
1975
P. dematus
4-8
2.4-2.8
7-9.5
-9.5
Smith and Fahay, 1970
P. olivaceus
5-6
2.6-2.8
7.1-8.7
10.2->14.2
Okiyama, 1967
Xystreurys liolepis
6
2.0
6.0-6.7
7.5->8.7
Original
Psciidorhomhus elevatus
9
_
5.5-6.4
-10
Devi, 1969
P. pentophlhalmus
7
—
7.1-7.6
8.7-12.2
Okiyama, 1974a; Minami, 1981a
Tarphops oligolepis
8
—
—
9.2-12.5
Okiyama, 1974a
Cithanchthys arctifrons
3
<2.3
-5-8
13-15
Richardson and Joseph, 1973
C. cornutus
3
<2.2
5.8-8.9
-18
Tucker, 1982
C. gymnorhinus
3
—
5.3-7.7
-18
Tucker, 1982
C. platophrys
3
<2.0
5.3-6.1
11.2-18.5
Original
C. sordidus
2
-2.0
10.4-11.4
20->39
Ahlstrom, 1965; original;
Ahlstrom and Moser, 1975
C. spiloplerus
2
—
5.7-6.8
9-12
Tucker, 1982
C. sligmaetis
0
-2.0
9.2-10.2
24.0->35.5
Ahlstrom, 1965; onginal;
Ahlstrom and Moser, 1975
Eiropus crossotus
2
<2.3
4.9-9.5
-11
Tucker, 1982
E. microslomus
0
<2.3
5-7
10-12
Richardson and Joseph, 1973
Cydopsella chillendeni
8-9
—
-7.5
>13.0
Evseenko, 1982a
C. fimbriala
~9
-1.5
6.9
14.0
Gutherz, 1970; Evseenko, 1982a
C. querna
8-11
—
—
>32
Ahlstrom, 1972a
Syacium guineensis
7
<2.1
<6.5
>13.9
Aboussouan, 1968b
S. gunleri
5-8
<1.8
—
—
Evseenko, 1982a
S. micruruin
5-8
<1.8
—
—
Evseenko, 1982a
S. ovale
5-8
-1.6
4.1-4.8
-14-- 20
Ahlstrom, 1972a; original
S. papillosum
5-8
<2.3
5.5-6.0
15-13
Futch and Hoff, 1971; Evseenko,
1982a
along the ventral midline, and small melanophores on the trunk,
tail, ventral gut, pectoral fin, brain and lower jaw. During flexion
the entire body except for the caudal fin base becomes solidly
pigmented, a darker band forms forward of the caudal peduncle,
and the snout becomes heavily pigmented.
Cilharuiae (Fig. 342). — Larvae of this family are known from
five specimens (4.0-8.0 mm) of Brachypleura novaezeelandiae
described by Pertseva-Ostroumova (1965). Notochord flexion
occurs between 5.0 and 7.0 mm and transformation at about
8.0 mm. Larvae have a moderately deep, thick body and a large
head with large jaws and eyes and about 10 large preopercular
spines; the sixth dorsal ray is elongate and the rays anterior to
it are assumed to be elongate, although damaged in all available
specimens; pelvic fins are elongate, extending beyond the anus;
pigment consists of a series of melanophores along the dorsum,
a series along the horizontal septum, and a postanal series along
the ventrum, melanophores below the gut, and on the pelvic
fin.
Scophthalmidae (¥'\g. 343). — Larvae are known for 9 of the 10
species of this family. Petersen ( 1 909) described 5 of the 7 species
occurring in the eastern North Atlantic; Jones (1972) provided
excellent illustrations of the 2 species of eastern Atlantic Scoph-
thalmus and Bigelow and Welsh (1925) described larvae of 5.
aquosus. the only western Atlantic representative of the family.
Newly hatched larvae are 2.0-4.0 mm in length (Table 171);
size at notochord flexion for most species is 6.0-8.0 mm. Meta-
morphosis can begin by 8 or 9 mm and be completed by 1 3
mm (S. aquosus. Phrynorhombus norvegicus, Zeugopterus
punctaius) or delayed to over 20 mm (S. maximus. S. rhombus).
Larvae are deep- and thick-bodied, especially at the gut, have
a large head and jaws and moderate to large eyes. Scophthalmid
larvae develop extensive head spination. Three species (Z. punc-
tatus, P. regius, Lepidorhoinbus whiffiagonis) develop paired
otic spines. In Z. punctatus, spines also develop at the lateral
aspect of the midbrain and on the opercle. Larvae of P. nor-
vegicus develop spines along the lower jaw, on the opercle and
preopercle, and at the shoulder (posttemporal region) while L.
boscii has preopercular spines and a shoulder cluster. S. ma.\-
imus and S. rhombus have a supraocular spiny ridge, numerous
spines on the opercle and preopercle and a shoulder cluster.
Pigmentation is heavy on the head and body in most species.
Z. pimctatus has a series of finfold bars and L. boscii develops
these and also incomplete bars on the body. Late larvae of all
species develop bars on the dorsal and anal fins.
Paralichthyidae (Figs. 344, 345). — Three subgroups are recog-
nized in this family on the basis of adult characters: Paralichlhys
and relatives (Ancylopsella. Gaslropsella. Hippoglossina. Lio-
glosslna. Verecundum. Xystreurys): Pseudorhombus and rela-
tives (Cephalop.tetta. Tarphops): and Cyclopsetta and relatives
(Cltharlchthys. Eiropus. Syacium).
In the first group larvae are known for species oi Paralichlhys
and Hippoglossina and for Xyslreurys liolepis and in the second
group larvae are known for Pseudorhombus and Tarphops. In
652
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 175. Meristic and Larval Characters of Bothidae.
Basip-
leryg-
2nd
Length al
Larvae
Urohyal
lal
Cleilhral
Otic
Scale
dorsal
transforma-
Taxon
described'
Number of vertebrae
spines-
spines
spines
spines
spines
ray'
tion (mm)
References
Taeniopsettinae
Engyophn's
+
10 + 27-31 =
= 37-41
+ +
+ +
+ +
+
0
M
-19
Hensley, 1977
Taeniopselta
+
10 + 30-32 =
= 40-42
+ +
+ +
+ +
+
0
s
-60
Amaoka, 1970
Trichopsetta
+
10-11 + 30-33 =
= 40-43
+ +
+ +
+ +
+
0
s
-28
Futch, 1977
Perissias
-
10 + 29-30 =
= 39-40
Bothinae
Arnogtossus
+
10-12 + 27-36 =
= 37-48
0
0
0
0
+
M, L
21-46
Kyle, 1913; Pertseva-
Ostroumova, 1965;
Amaoka. 1973, 1974
Bolhus
+
10 + 25-32 =
= 35-42
0
0
0
0
0
M, L
9-42
Kyle, 1913
Chascanopselta
+
16-18 + 37-44 =
= 53-59
0
0
0
0
0
M
78-120
Bruun, 1937; Nielsen,
1963b; Amaoka, 1971;
Pertseva-Ostroumova,
1971
Crossorhomhus
+
10 + 24-27 =
= 34-37
0
+
0
0
+
M, L
15-20
Ochiai and Amaoka, 1963
Engyprosopon
+
10 + 23-27 =
= 33-37
+ +
+ +
0. + +
0
0
M
16-18
Pertseva-Ostroumova,
1965
Grammalobolhus
+
10 + 27-28 =
= 37-38
+
+
0
0
0
M
-15
Kamoharaia
+
13-14 + 37-39 =
= 50-53
0
0
0
0
0
M
-91
Nielsen. 1963c
Laeops
+
10-12 + 35-42 =
= 46-53
0
0
0
0
0
L
51-80
Balaknshnan, 1963;
Amaoka, 1972a; Hubbs
and Chu, 1934
Lophonectes
+
10 + 32-33 =
= 42-43
0
+
0
0
0
M
-20
Original
Monolene
+
10 + 28-38 =
= 38-48
0
0
0
0
0
L
-30
Futch, 1971
Pelecanichlhys
+
17 + 40 =
= 57
0
0
0
0
0
M
>90
Struhsaker, pers. comm.
Pseltina
+
10 + 29-30 =
= 35-40
+
+
0
0
+
M
16-20
Pertseva-Ostroumova.
1965; Amaoka. 1976
Asterorhombus
Japonotaeops
Mancopsetta
Neolaeops
Parabolhus
Tosarhombus
10 + 26-27 = 36-37
11 + 41-44 = 52-53
13-16 + 38-50 = 52-66
13 -I- 38 = 51
10-1- 31-36 = 41-46
10 -I- 28-30 = 38-40
= yes;
^ 0 = absent; + = present; + + = strong.
^ S = shon; M = moderate; L - long.
these two subgroups hatching, notochord flexion and meta-
morphosis occur at a small size (Table 1 74). Larvae of these
groups are noted for a dorsal crest consisting of elongate early
forming rays, beginning with the second dorsal ray (Table 1 74).
Larvae of the Paralichthys group are moderate in body depth,
with a deep head and moderate-size jaws. Body thickness is
moderate except that Paralichthys is more laterally compressed
than in other genera reported (Fig. 344). The gut mass is large.
Preopercular spination consists of an anterior and posterior se-
ries in Paralichthys. Pseudorhombus and Tarphops and an an-
terior series only in Hippoglossina. Larvae of Paralichthys den-
talus have one to several minute sphenotic spines (Smith and
Fahay. 1970) and P. olivaceus develops a spine cluster on the
sphenotic, one spine on the epiotic. and 1-2 spines on each bone
in the opercular series. Larvae of Pseudorhombus pentophthal-
inus have a single sphenotic spine, and some on the opercular
bones (Okiyama, 1 974a); Devi ( 1 969) shows two rows of sphen-
otic spines in P. elevatus.
Yolk-sac larveofthe Paralichthys and Pseudorhombus e,Toups
develop moderate to heavy pigmentation with some on the
finfolds. Later-stage larvae have pigment over the brain, on the
lower head and jaw region and below and lateral to the gut.
Most species have a melanophore series along the dorsum and
ventrum. Lateral pigment may consist of a series along the
horizontal septum {Paralichthys. Tarphops), a wide-spread zone
of melanophores (Xysireurys. Hippoglossina) or a posterior bar
(Pseudorhombus penlophlhalmus). Most species have a series
of internal melanophores above the spinal column and some
melanophores on the posterior region of the finfold and devel-
oping dorsal and anal fins.
Larvae of the Cyclopsetta assemblage are similar morpholog-
ically to those of the Paralichthys and Pseudorhombus assem-
blages, but differ in spination and fin ray development. The rays
forming the dorsal crest are typically longer and stand out more
abruptly compared with Paralichthys and associated genera. The
fin ray complement of the crest, along with other characters,
divides the assemblage into two generic pairs: Citharichthys-
Etropus and Cyclopsetta-Syacium. Species of the former group
have either two or three elongate rays, except for two species
which lack a crest altogether (Table 1 74). Species of Syacium
have 5-8 elongate dorsal rays and 8-1 1 occur in Cyclopsetta.
The left pelvic fin forms before the right and may develop elon-
gate rays in some species. The first two pelvic rays are elongate
in Citharichthys sordidus and C. plalophrys. the second ray only
is elongate in C. cornutus. C. gymnorhinus. C. spilopterus and
Etropus crossotus; C. arctifrons. C. stigmaeus and E. micros-
AHLSTROM ET AL.; PLEURONECTIFORMES
653
Fig. 346. Larvae of Bothidae. (A) Trichopselta vcntralis. 21.9 mm, from Evseenko. 1982a; (B) Engyophrys senla. 12.3 mm. from Hensley.
1977; (C) Taeniopsetla oceltala. 59.0 mm, from Amaoka, 1970; (D) Monolene sessiticauda. 14.3 mm, redrawn from Futch, 1971; (E) Psettina
hamancnsis. 4.2 mm, from Pertseva-Ostroumova, 1965; (F) P. hamanensis. 18.1 mm. ibid.
loniits lack elongate pelvic rays. The first three pelvic rays be-
come markedly elongate in Cyclopsetta and the entire left fin
becomes moderately elongate in Syacium.
Eiropus and Citharichthys (except for C. arctifrons) develop
one or more rows of small preopercular spines. According to
Tucker (1982), small frontal-sphenotic spines are present in
some species of Cithanchlhys and Etropus (6-8 spines on each
side in C. conmtus. up to 6 in C. gymnorhinus. 1-2 m C. spi-
lopterus, and 3-4 in E. crossotus). Syacium and Cyclopsetta
develop a series of large preopercular spines at the margin of
the bone and, in some species, an irregular anterior series. The
spine at the angle of the primary series becomes antler-like in
preflexion larvae of Syacium and in postflexion larvae of Cy-
clopsetta. Early preflexion larvae of Syacium develop single
elongate sphenotic spines which remain prominent during the
remainder of the larval period. Sphenotic spines in Cyclopsetta
are early-forming but short.
Larvae of both subgroups of the Cyclopsetta assemblage typ-
ically have pigment above the brain, on the lower head region,
below the gut, lateral to the posterior region of the gut, and
above the gas bladder. Early preflexion larvae of most species
have a series of small postanal melanophores and a bar or a
654
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 76. Size Data for Pleuronectinae Larvae.
Size al
hatching (mm)
Size al noto-
chord flexion
(mm)
Size at
transformation
(mm)
Acanthopsetta nadeshnyi
Athereslhes evennanni
A. slomias
Cteisthenes herzensleini
Embassichlhys halhyhius
Eopsella grigorjewi
E. jordani
Glyptocephalus cynoglossus
G. stelleri
G. zachirus
Hippoglossoides dubius
H. elassodon
H. platessoides
H. robustus
Hippoglossus hippoglossus
H. slenolepis
Hypsopsetta gultutata
Isopsella isolepis
Kareius hicoloralus
Lepidopsella hilineata
L. mochigarei
Limanda aspera
L. ferrugmea
L. limanda
L. punclalissima
L. schrenki
L. schrenki (as Pseudopleuronectes
yokohamae)
L. yokohamae
Liopsetla glacialis
L. obscura
L. pinnifasciata
L. pulnami
Lyopsella exilis
Microstomiis achne
M. kill
M. pacificus
Parophrys velulus
Plalichlhys flesus
P. stellatus
P. pallasii (as Platessa quadriluberculata)
P. platessa
Pleuronichthys coenosus
P. cornutus
P. decurrens
P. rilien
P. verlicalis
Psetlichlhys melanostictus
Pseudopleuronectes americanus
P. herzensteini
Reinhardtius hippoglossoides
Tanakius kitaharai
Verasper variegatus
<3.0
8.4-9.9
ca. 20-?
<8.4
ca. 11.5-15
—
—
—
32.9-?
2.2-2.6
7.0-7.9
8.1-11.4
ca. 9.0
_
1 6.2-7
2.5-3.0
7.2-8.9
11.4
2.8
_
_
3.5-5.6
15-21
25-?
4.1-5.2
15-17
19-48
ca. 6
15.3-24.0
49-72
3.0-3.4 TL
<12.4
18.1
5.4-6.6 TL
9.0-10.2
—
4.0-6.0
9.5-17.5
18-30
ca. 4.0
ca. I1-?
>28.6
6.5-7.0
16-18
22-34
7.8-8.5
13.6-17.8
14.7-24.1
L7-2.3
4.0-5.2
4.4-:>8.8
2.7-2.9
9.1-14.0
15->21.9
ca. 3.0
ca. 4.0-9.0
ca. 14-?
3.4-3.8
ca. 8.4-9.9
>17.7
3.95-4.48
ca. 8.9
_
2.2-2.8
7.5-9.5
ca. 10-?
2.0-3.5
5.9-ca. 10
ca. 14
2.7-4.0
7-8.7
12-20
1.79-2.21
—
8.1->9.6
ca. 2.4
_
_
ca. 2.4
<7.4
1 2.0-?
3.5-3.8
ca. 7.0
ca. 7.5-10.0
3.7
—
—
2.5-3.5
-6.6
>9.0
3.15-3.93
8.11-8.45
>8.5
3.1-3.6
6.0-7.1
7.3
ca. 5.6
9.0-10.9
15.7-24.7
_
8.8
—
4.84
12-15
18-28
ca. 6.0
ca. 10-15
ca. 20->45
2.3-2.8
8.8-10.5
ca. 20
2.25
5.9-7.1
9-12
1.9-2.1
5.5-6.0
-
5.6
8.9
ca. 10.0-?
6.0-7.5
8.9-10.2
10.5-14
3.9
6.2-8.5
8.2->11.4
2.65-2.8
>3.6
7.25-13.0
4.9-5.5
7.8-11.0
10.5->21.0
2.1
4.3-5.6
6.0->10.0
2.4
5.0-7.2
7.3->11.0
<3.0
ca. 8.0
>22.6
2.3-3.5
5.0-7.1
6.8
2.6-2.9
ca. 6.0-8.5
ca. 10.4-?
10-16
25-27
45-65
ca. 3.0
—
18.9-ca. 20
3.8
ca. 9-12.4
ca. 16.4-?
Pertseva-Ostroumova, 1961
Pertseva-Ostroumova, 1961
Pertseva-Ostroumova. 1961
Okiyama and Takahashi, 1976; Dekhnik, 1959
Richardson, 1981b
Okiyama and Takahashi, 1976
Alderdice and Forrester, 1971
Petersen, 1904
Okiyama, 1963; Dekhnik, 1959; Pertseva-
Ostroumova, 1961
Original; Ahlstrom and Moser, 1975
Okiyama and Takahashi, 1976
Dekhnik, 1959; Pertseva-Ostroumova, 1961
Petersen, 1904; Russell, 1976; Nichols,
1971; Colton and Marak, 1969
Pertseva-Ostroumova, 1961
Schmidt, 1904; Russell, 1976 (summary)
Thompson and Van Cleve, 1936; Pertseva-
Ostroumova, 1961
Sumidaet al., 1979
Richardson et al., 1980
Pertseva-Ostroumova, 1961
Pertseva-Ostroumova, 1961
Yusa, 1958; Okiyama and Takahashi, 1976
Dekhnik, 1959; Pertseva-Ostroumova, 1961
Bigelow and Welsh, 1925; Miller, 1958
Russell, 1976 (summary)
Pertseva-Ostroumova, 1961
Hikila, 1952
Pertseva-Ostroumova, 1961
Yusa, et al., 1971; Minami, 1981a
Pertseva-Ostroumova, 1961
Pertseva-Ostroumova, 1961; Kurata, 1956
Pertseva-Ostroumova, 1961
Laroche, 1981
Original; Ahlstrom and Moser, 1975
Okiyama and Takahashi, 1976
Petersen, 1904
Original; Ahlstrom and Moser, 1975
Onginal; Budd, 1940; Ahlstrom and Moser, 1975
Nichols, 1971; Russell, 1976 (summary)
Orcutt, 1950; Yusa, 1957; Pertseva-
Ostroumova, 1961
Pertseva-Ostroumova, 1961
Nichols, 1971; Russell. 1976 (summary)
Sumidaet al., 1979; Budd, 1940
Takita and Fujita, 1964; Minami, 1982a
Sumida et al., 1979; Budd, 1940
Sumida et al., 1979
Sumida et al., 1979; Ahlstrom and Moser, 1975;
Budd, 1940
Hickman, 1959
Breder, 1923; Laroche, 1981
Dekhnik. 1959
Jensen, 1935
Okiyama and Takahashi, 1976
Takita et al., 1967; Pertseva-Ostroumova,
1961; Uchida, 1933
shorn lateral pigment series posteriad on the tail. In some species
of Cithanchthys the ventral series coalesces into a more sparse
series of larger spots and a similar series develops along the
dorsum (e.g., C. arctiforns. C. cornutus. C. sordidus). In other
species, series along the dorsum and ventrum are abbreviated
or absent and only the tail bar may be present (e.g., C. gym-
norhinus. C. platophrys) or absent (C. spilopterus). Etropus lar-
vae have dorsal and ventral series and cither a short lateral
series {E. crossotus) or a long one (C inicroslomiis). Cyclopsctta
and Syacnim have dorsal and ventral series and a short lateral
series posteriad on the tail. Fin pigment is principally on the
spatulate tips of the elongate dorsal and pelvic fin rays. Late-
Fig. 347. Larvae and transforming specimens of Bothidae. (A) Crossorhombus kobensis, 16.0 mm, from Amaoka, 1979; (B) Engyprosopon
xenandrus. ca. 20.0 mm; (C) Lophonectes gallus. 18.5 mm, original. K 1 38/74, New Zealand; (D) Bothits thompsoni. ca. 36.0 mm; (E) B. mancus.
ca. 30.0 mm. B, D, and E from P. Struhsaker, unpublished.
656
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 348. Larvae of Bothidae. (A) Arnoglossus dehilis. ca. 59.0 mm, from P. Struhsaker, unpublished; (B) Chascanopsetta lugiihris. 1 20.0 mm,
from Amaoka, 1971; (C) Laeops kilaharae. 79.0 mm. from Amaoka, 1972; (D) Pelecanichthys sp., ca. 95.0 mm. from P. Struhasker, unpublished.
AHLSTROM ET AL.: PLEURONECTIFORMES
657
Table 177. Size Data for Larvae of Achirinae and Soleinae.
Si7c al
Si/e al
hatching
Size at
Iransformalion
Taxon
(mm)
flexion (mm)
(mm)
References
Achirinae
Achirus linealus
1.6
ca. 2.5-3
3.0-5.5
Houdeetal., 1970
Trinecles maculatiis
1.7-1.9
ca. 3-4
ca. 5
HiWebrand and Cable, 1938
Soleinae
Aesopia cornula
4.1
—
9.2-?
Mito, 1963
Austroglossus nucrolepis
1.7
5.5-6.6
8.5-ca. 16
OToole. 1977; Brownell. 1979
Balhysolea profundicola
—
—
4.32
Aboussouan, 1972c
Buglossidium lutcurn
1.8-2.3
ca. 6-8
ca. 6-10
Holt, 1891; Ehrenbaum, 1897
Dicologoglossa cuneala
1.3
ca. 6.3-6.5
7-7.5
Lagardere, 1980; Lagardere and Aboussouan, 1981
Euryglossa pan
<2.66
<4.6
3.4-6.5
Jones and Menon, 1951
Heteromyctens capensis
1.7
?-6.5
ca. 6.2-?
Brownell, 1979
H. japonicus
—
ca. 4.55
5.0-7.0
Minami, 1981b
Microchirus hoscanion
—
—
7.2
Aboussouan, 1972c
M. frechkopi
—
—
5.68
Aboussouan, 1972c
M. ocellatus
2.0
4.6-5.1
6.8->8.2
Palomera and Rubies, 1977
M. vanegalus
2.4-2.9
6.1->7.1
ca. 7-12(18)
Cunningham, 1890; Petersen, 1909
Pegusa cadenali
—
—
7.0
Aboussouan, 1972c
P. impar
—
—
8.5->12
Padoa, 1956k
P. lascaris lascaris
<3.5
5.3-8.1
9.5->11.2
Clark, 1914
P. lascaris nasuta
2.1-2.5
_
_
Dekhnik, 1973; Padoa, 1956k
Solea cuneala
—
—
7.0
Aboussouan, 1972c
S. heinti
<2.2
>2.7-3.2
_
Balakrishnan and Devi, 1974
S. hexophthalma
—
—
8.0
.Aboussouan, 1972c
S. ovaia
—
ca. 3-4
4. 5-''
Balaknshnan, 1963
S. solea
2.5-3.8
5.5-?
ca. 7-14.6
Russell, 1976 (summary)
Synaplura kleini
3.0
?-6.5
ca. 7-9
Brownell, 1979
Zehrias japonicus
4.1 TL
—
—
Mito, 1963
Z. zebra
4.0 TL
-
-
Mito, 1963
stage larvae of most species develop chevron-shaped bars on
the epaxial and hypaxial myosepta. Metamorphosing specimens
of Cyclopsetta have series of large ocelli on the dorsal and anal
fins. Dorsopsetta norma described by Nielsen (1963b) on the
basis of two metamorphosing specimens is apparently a species
of Cyclopsetta.
Bolhidae (Figs. 346-348). — Two bothid subfamilies are recog-
nized, Taeniopsettinae and Bothinae. Bothid larvae are thin-
bodied to diaphanous, sparsely pigmented, and all develop an
elongate second dorsal ray (Table 1 75). Also, spines may appear
on the urohyal, basipterygia, cleithra and epiotics in a pattern
which is generally consistent for subfamilies and genera (Table
175). Bothid larvae reach a relatively large size before meta-
morphosis. Early larval stages are often poorly represented in
collections.
Larval series are known for all taeniopsettine genera, except
Perissias. Larvae of Tnchopseita and Engyophrys are ovate while
those of Taemopsetta are round (Fig. 346). All have a complete
complement of head spines (Table 1 75). The second dorsal fin
ray is slightly or moderately elongate. Taeniopsetta lacks me-
lanophores, but live larvae have four reddish-orange spots along
the bases of the dorsal and anal fins, and orange, reddish and
yellow blotches and bands on the body and head. Tnchopsclla
has three series of melanistic blotches along the dorsal and anal
pterygiophores and along the body axis (left side). Engyophrys
lacks melanophores.
Larvae of Bothinae have an ovate, round, or elongate shape
(Figs. 347, 348) and lack epiotic spines. Engyprosopon has nu-
merous urohyal and basipterygial spines and some species have
numerous spines on the cleithrum. Psellina and Grammato-
hothiis have urohyal and basipterygial spines, and early larvae
of the former have a hook-like projection on the lower jaw (Fig.
346). Crossorhombiis and Lophonectes have basipterygial spines
only and all other known bothid larvae lack head spines. Cros-
sorhombus larvae have a series of scale spines along the bases
of the dorsal and anal fins, one scale per ray, and species of
Pseltina and .-irnoglossus also develop such scale spines. In the
species of .•lr«(),^/aM(« described by Kyle (1913), patches of scale
spines develop on the median and ventral regions of the ab-
domen. The second dorsal ray is usually moderately elongate
but can be greatly elongate and ornamented, as in Arnoglossus.
Pigmentation is sparse in most bothine larvae and lacking in
some species. Exceptions are found in species of .Arnoglossus
and Psettina which usually have melanophores above the brain,
ventrally on the gut, above the gas bladder, in series along the
dorsal and ventral midlines, and along the horizontal septum;
in some species a complete or partial bar is present posleriad
on the tail. Preflexion larvae of Bothus have a melanistic blotch
near the tip of the notochord; later larval stages are unpig-
mented, except that transforming specimens of some species,
B. myriaster (Amaoka, 1964) and B. mancus(Fig. 347), become
heavily spotted over the body and fins, Laeops has melanistic
blotches forming an irregular pattern over the body and median
fins.
Monolene shares some adult characters with taeniopsettines
but larval characters place it with the bothines. Larvae are elon-
gate, lack head spines, have an elongate ornamented second
dorsal ray, and melanistic pigment above the gut, on the right
side of the brain and on the dorsal fin membrane (Fig. 346).
Pleuronectidae (Figs. 349-355).— Of the five pleuronectid
658
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
AHLSTROM ET AL.: PLEURONECTIFORMES
659
subfamilies recognized (see introduction), the Pleuronectinae is
the largest with 26 genera, representing % of the genera in the
family. Three contributions that summarize egg and larval in-
formation for pleuronectine flatfishes from the eastern North
Atlantic and Mediterranean are Ehrenbaum (1905-1909), Pa-
doa (1956k). and Nichols (1971). Bigelow and Welsh (1925).
Bigelow and Schroeder (1953), Martin and Drewry (1978), and
Fahay (1983) give information on eggs and larvae of western
Atlantic pleuronectine flatfishes. The most comprehensive work
dealing with early life history stages of flatfishes from the western
North Pacific is Pertseva-Ostroumova (1961).
Yolk-sac larvae of pleuronectine flatfishes can be as small as
1.7 mm (Hypsopseila giittulata) or as large as 10-16 mm (Rein-
hardtius hippoglossoides) and size at hatching is a primary char-
acter for identifying yolk-sac larvae (Table 1 76). The pigment
pattern can be quite distinctive, as for example in the genus
Pleuronichthys. but in many pleuronectines the body pigment
migrates during the yolk-sac stage, and is variable from speci-
men to specimen of the same species. The yolk sac itself can
lack pigment (as in Parophrys vetulus. Hippoglosstis stenolepis
or Eopsetta jordani), can be moderately pigmented (as in Lyop-
setta exilis, Lepidopsetta bilineata or Psettichthys melanostictus)
or can be heavily pigmented (as in Pleuronichthys decurrens or
I'erasper variegaliis). Similarly the finfold can lack pigment or
be variously pigmented and useful in identification.
Early preflexion pleuronectine larvae are slender; the head is
of moderate size; snout-anus length can be as much as 50% NL
(as in four species oi Pleuronichthys larvae, Sumida et al., 1979)
but usually is shorter (i.e., 35-45% NL). The gut is initially
straight but develops a coil soon after the completion of yolk
absorption. Greatest body depth after the gut becomes looped
is either at the anus or slightly anterior to it. Body shape of
preflexion larvae is quite similar from species to species. There
are few distinctive characters unique to the larval period of
pleuronectine flatfishes. Only a few kinds of pleuronectine larvae
develop head spination. Preopercular spines form in larvae of
Athercslhcs. Glyptocephalus, Tanakiiis and E opsei! a; oiic spines
develop on larvae of Microstomus (at least on 2 species), Hyp-
sopsetta, and Pleuronichthys ( 1 species); Athercsthes has a spi-
nous supraocular crest. Head spination develops during the pre-
flexion stage, but usually is best developed on flexion or early
postflexion larvae.
The caudal fin begins forming either slightly before or together
with the dorsal and anal fins. The first caudal supporting bones
to form as cartilage are the hypurals. Usually several caudal rays
(2 + 2 or 3 + 3) are formed before flexion begins. In late flexion
and early postflexion larvae, the end of the notochord can project
beyond the hypural plates. The complete complement of caudal
rays is usually laid down during the flexion period.
The dorsal and anal fins form in the finfold at some distance
from the main part of the body. The intervening space becomes
filled with the pterygiophores that support the dorsal and anal
fin rays, causing an increase in body depth. In both dorsal and
anal fins the rays begin forming at the anterior ends of the fins
and the differentiation proceeds posteriad. The first few rays in
both fins are reduced in size and the terminal ray is often minute.
Pelvic fin buds usually form during the flexion stage but pelvic
rays usually are not developed until the postflexion stage. As in
all flatfishes, formation of pectoral fin rays is delayed to the end
of the transformation stage.
The vertebral processes ossify before the centra. In the caudal
group of vertebrae, ossification of haemal and neural processes
proceeds posteriad. Ossification of abdominal neural processes
can follow several patterns, but usually proceeds anteriad. The
last neural and haemal processes to ossify are the truncate spines
of the 2 or 3 vertebrae anterior to the urostyle. Centra ossify
initially at the bases of neural and haemal processes and ossi-
fication proceeds peripherally until a complete ring is formed.
On first formation only the middle portion of a vertebral cen-
trum is ossified, hence the space between adjacent centra may
be as wide as the ossified portions of the centra. The ural centra
are the first to ossify in some pleuronectines or they can ossify
at the same time as other centra. The last centra to form are
those of the 2 (or 3) vertebrae anterior to the urostyle.
All pleuronectine larvae that have been described have body
pigment. The pigment pattern changes with growth, often mark-
edly. Also, there is often considerable variation in pigmentation
of larvae of similar sizes of the same species. Notwithstanding,
body and finfold pigment constitutes a primary character for
identification of flatfish larvae during the preflexion stage.
To show the variety of pigment patterns found on preflexion
stage pleuronectine larvae, preflexion larvae of 1 7 species from
the North Pacific are illustrated (Figs. 349-351). Heavily pig-
mented larvae are in the genera Pleuronichthys. Hypsopsetta,
and I'erasper (Fig. 349). The posterior portion of the tail is
unpigmented or pigment is confined to marginal spots along the
notochord. The unpigmented tail area is more extensive in some
species than in others. Finfold pigment is very useful in iden-
tifying these larvae to species taken in conjunction with larval
size and extent of tail pigment.
In the other 14 kinds of larvae representing as many genera,
tail pigment appears in a number of patterns. The larvae illus-
trated in Figs. 350 and 35 1 are arranged in the order of increasing
complexity. In the simplest pattern pigment is concentrated
along the ventral midline with only moderate dorsal or lateral
pigment, as in Hippoglossus stenolepis or Reinhardtius hippo-
glossoides. Although Parophrys vetulus and Lyopsetta exilis have
more ventral margin pigment than dorsal, it is almost contin-
uous on both margins. Platichthys stellatus has more diffused
pigment over the tail portion of the body, but it is not in a
pattern. The most unusual pigment is found in Atherestes. There
are two conspicuous dorsal patches as opposed to almost no
ventral pigment. Pigment on Eopsetta jordani is limited to a
mid-tail band and a terminal notochord patch. A more common
pattern is encountered in Isopsetta, which has two pigment bands
across the tail together with the terminal notochord pigment. A
basically similar pattern is found in Lepidopsetta bilineata. Pset-
tichthys is unusual in having alternating dorsal and ventral
blotches. Hippoglossoides elassodon has three tail pigment areas
(i.e., opposing dorsal and ventral pigment patches) together with
terminal notochord pigment. This is also the basic pattern in
Microstomus. Emhassichthys increases opposing tail patches to
Fig. 349. Larvae and transforming specimens of Pleuronectidae. (A) Pleuronichthys coenosus. 3.7 mm, from Sumida et al., 1979; (B) P.
coenosus. 8.9 mm. ibid; (C) Hypsopseila giitndala. 2.6 mm, ibid; (D) //. giittulata. 6.6 mm, ibid; (E) I'erasper variegaliis. 5.6 mm, from Pertseva-
Ostroumova. 1961 after Uchida, 1933; (¥) \'. variegatus. 12.4 mm. ibid.
660
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
four plus terminal notochord pigment (Richardson, 1981b);
G/yptocephalus zachirus has pigment bands which alternate with
ventral patches, plus the terminal notochord pigment.
At least four other genera of pleuronectine flatfishes occur in
the eastern North Pacific. The preflexion stage larvae of Pleu-
ronectes pallasii, Liopsetta glacialis. and Limanda aspera lack
melanistic bands (Pertseva-Ostroumova, 1961). Larvae are un-
known for the fourth genus, Clidoderma.
Larvae are known for species representing six additional gen-
era in the western North Pacific. According to Pertseva-Os-
troumova (1961), preflexion larvae o{ Acanthopsetta nadeshnyi
and Kareius bicoloratiis lack bands; those of Cleisthenes her-
zensteini (see also Okiyama and Takahashi, 1976), Pseudopleu-
ronectes herzensteini and P. yokohamae (see also Dekhnik, 1 959;
Yusa, 1960a, b; Yusa et al., 1971) have two tail pigment bands
plus terminal notochord pigment. Preflexion larvae of I erasper
vahegatus (Fig. 349) are as heavily pigmented as those of Pleii-
ronichthys (Takita et al., 1967; Uchida, 1933). The pigment
pattern on preflexion larvae of Tanakius kitaharai is very sim-
ilar to that on larvae of Glyplocephalus stelleh (Okiyama and
Takahashi, 1976). Larvae have not been described for the mono-
typic genus Dexistes.
In species with banded preflexion larvae, the bands usually
persist into later larval stages; those with diffuse or linear pig-
ment patterns generally do not develop bands in later stages,
although pigment may become associated with myosepta (Figs.
352, 353). Virtually all late postflexion and metamorphic pleu-
ronectines develop a distinct pattern of bars or blotches on the
body and median fins, which persists into the juvenile stage
(Fig. 354).
Of the four other pleuronectid subfamilies, larvae have not
been described for Paralichthodinae, while some information is
available on the Samarinae, Poecilopsettinae, and Rhomboso-
leinae. Pertseva-Ostroumova (1965) described two larval spec-
imens (6.4, 8.7 mm) of Samaris cnslatus and Struhsaker (pers.
comm.) has described large pelagic larvae of Samariscus sp. and
Poecilopsetta hawaiiensis (Fig. 355). Larvae of S. cristatus are
deep-bodied in the gut region, have a relatively large head and
jaws and a pigment pattern consisting of melanophore patches
along the dorsum and ventrum. along the outer margins of the
pterygiophore zones, and along the dorsal and anal fins; the
ventral region of the gut is pigmented. A series of Samariscus
triocellatus, 7.3-19.0 mm (provided by Dr. T. A. Clarke, Univ.
of Hawaii), is similar to Samaris cristatus in having a slender
body and wide pterygiophore zones but the gut coil is elongate,
protrudes beyond the ventral profile, and the fourth dorsal ray
is elongate. The left eye has begun to migrate at 7.3 mm and is
at the dorsal midline by 12.0 mm. Larvae of Samariscus cor-
allimts are similar but attain a larger size (ca. 26 mm). Both
species lack pigment. Late postflexion larvae of Poecilopsetta
have a body form similar to samarines (slender body with wide
pterygiophore /ones) but have a different gut structure, no elon-
gate dorsal ray, and have a striking pigment pattern consisting
of dorsal and ventral myoseptal series and large blotches over
the pterygiophore zones, dorsal and anal fins, and gut (Fig. 355).
A 29-mm late postflexion larva from the North Atlantic has a
pigment pattern identical to Hawaiian specimens.
Reared yolk-sac and early preflexion larvae of rhombosoleine
species have been illustrated and briefly described: Ammotretis
rostratus (Thomson, 1906); Rhomhosolea plebeia (Anderton,
1907); Colistnim glint hen. Pclotretis flavilatus. and Peltorham-
phus novaezeclandiae (Thomson and Anderton, 1921). The oil
globules remain evenly dispersed throughout the yolk-sac pe-
riod. Heavy melanistic pigmentation develops on the head, body,
yolk sac, and finfold. Late yolk-sac larvae of C gunthcri deve\op
an unusual lobate projection of the dorsal finfold, which extends
well anterior to the head. A similar structure appears in yolk-
sac larvae of the soleid, Pcgusa lascaris (Holt, 1891). Rapson
( 1 940) described and illustrated with photographs a reared series
of Pelot ret is Jlavilat us. Flexion-stage larvae of this species are
deep-bodied and similar in appearance to paralichthyids, al-
though they lack elongate dorsal fin rays (Fig. 355). Pigmenta-
tion consists of dorsal and ventral midline series, series above
and below the spinal column, a linear patch below the gut, and
embedded melanophores in the otic region. Postflexion larvae
become mottled with large blotches on the body and fins. Cross-
land (1981) briefly described and illustrated pre- and postflexion
stages of a similar larva which he identified as Pe/torhamphus
latus and stated that Rapson's ( 1 940) series was a species of
Peltorhamphus. Crossland's (1982) illustration of a flexion-stage
Pelotrelis flavilatus has heavy pigmentation, a protruding gut
mass and looks very much like a soleid.
Soleidae (Fig. 356).— Two subfamilies, Soleinae and Achirinae,
are recognized in the family. In the Soleinae, life history stages
are well known for the eastern North Atlantic species, Solea
solea. Microchirus varicgaius. Buglossidium luteum and Pcgusa
lascaris (references summarized in Ehrenbaum, 1905-1909 and
Russell, 1976). A comprehensive volume on the development
of 5. solea was produced by Fabre-Domergue and Bietrix (1905).
Padoa (1956k) summarized information on eggs and larvae of
soles from the Mediterranean, and Aboussouan (1972c) briefly
Fig. 350. Larvae of Pleuronectidae. (A) Hippoglossus stenolepis. 1 5.0 mm, from Pertseva-Ostroumova. 1 96 1 ; (B) Reinhardlius hippoglossoides.
17.0 mm, from Jensen, 1935; (C) Lyopselta exilis. 5.9 mm from Ahlstrom and Moser, 1975; (D) Parophrys vetulus. 4.3 mm, ibid; (E) Ptatichthys
slellatus. 2.6 mm, from Orcutt, 1950; (F) Atheresthes stomias. 10.5 mm, original; (G) Eopsetta jordani. 6.2 mm, from Alderdice and Forrester,
1971.
Fig. 351. Larvae of Pleuronectidae. (A) Isnpsetta isolepis. 9.5 mm, original, CalCOFI 7205, Sta. 40.38; (B) Lepidopsena bihneata. 4.6 mm,
from Pertseva-Ostroumova, 1965; (C) Pseltichlhys melanoslictus. 6.7 mm, original, CalCOFI 5807 Sta. 40.38; (D) Hippoglossnide.s eta.ssodon.
9.2 mm, from Pertseva-Ostroumova, 1961; (E) Microstornus pacificus. 7.0 mm, redrawn from Ahlstrom and Moser, 1975; (F) Embassichlhys
balhybius. 18.5 mm, original, CalCOFI 4905, Sta. 29.83; (G) Glyplocephalus zachirus. 22.8 mm, redrawn from Ahlstrom and Moser, 1975.
Fig. 352. Larvae of Pleuronectidae. (A) Lyopselta exilis. 14.7 mm, original. CalCOFI 7805, Sta. 100.29; (B) Parophrys vetulus. 16.0 mm,
redrawn from Ahlstrom and Moser, 1975; (C) Isopsetla isolepis. 14.2 mm, original, CalCOFI 7205, Sta. 40.38, (D) Eopsetta grigorjewi, lO.O mm,
from Okiyama and Takahashi, 1976; (E) Pseltichlhys melanoslictus. 9.4 mm, original, CalCOFI.
AHLSTROM ET AL.: PLEURONECTIFORMES
661
B
D
G
662
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
B
D
AHLSTROM ET AL.: PLEURONECTIFORMES
663
f^ ,--V , * .
1 1 :^i;iiri'<<'*>ii:<>^<;^i^^ :
■ "i^r.pr-- •
664
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 178. Size Data and Number of Elongate Dorsal Rays for Larvae of Symphurinae and Cynoglossinae.
Number of
Size al
elongate
hatching
size al flexion
Size al trans-
Taxon
dorsal rays
(mm)
(mm)
formalion (mm)
References
Symphurinae
Svmphuriis atricaiida
5
1.9
9.4-10.8
19-24.2
Onginal
S. lacleus
4-5
—
6-<10
18
Kyle, 1913
S. ligulala
5
—
<10.5
32
Kyle, 1913
S. orienlalis
5
—
>4.4-<9.3
>12.3
Pertseva-Ostroumova, 1965
S. ptagiusa
4-5
<1.3
<6.2
<13
Hildebrand and Cable, 1930;
Olney and Grant, 1976
Cynoglossinae
Cynogtossus ahbrevialus
(as Areliscus trigrammus
3.2
Fujita and Takila, 1965
C. arel (as C. otigolepis)
2
—
—
—
Pertseva-Ostroumova, 1965
C. btlineatus
2
—
>4.0
6-8
Vijayaraghavan, 1957;
Pertseva-Ostroumova, 1965
C. brevis
—
—
—
ca. 4.0
Balakrishnan and Devi, 1974
C. capensis
2-A
1.2
ca. 9.9
10-15
Brownell, 1979
C. cynogtossus
2
<1.6
ca. 4. 1
<4.7
Balakrishnan and Devi, 1974
C. kopsi (as C. sibogae)
2
—
7.6
—
Pertseva-Ostroumova, 1965
C. lida
2
<2.1
>4.6
—
Balakrishnan and Devi, 1974
C. lingua
—
_
—
17.7-?
Jones and Menon, 1951
C. macroslomus
2
<2.5
4.2
ca. 5
Seshappa and Bhimachar, 1955
(as C. semifasciatus)
C. monopus
2
—
5-7
7.0
Balakrishnan, 1963
C. puncliceps
2
<1.4
ca. 4.2-?
ca. 4-5
Balakrishnan and Devi, 1974
C. robustus
2
1.85
—
_
Fujitaand Uchida, 1957
C. semifasciatus
2
<2.0
7.2-11
11-12.5
Balaknshnan, 1961
Paraplagusia japonica
2
-
<10.2
ca. 12.2
Minami, 1982b
described several species from off west Africa. Life history seires
have been described for two achirine species of the western
North Atlantic, Trinectes maculatus and Achirus lineatus. Eggs
of achirines are smaller than in most soleines (Table 172) and,
accordingly, size at hatching is also smaller; achirines and some
soleines undergo notochord flexion and transformation at vei^
small sizes (Table 1 77). Achirines are deep-bodied, with a large
gut that occupies a major portion of the body volume, a large
deep head with a distinct dorsal hump; eyes and jaws are large
(Fig. 356). Preflexion larvae of.-J. //«ea/M5 develop spinous ridges
above the eye (frontal bone), at the otic region (parietal and
autopterotic bones) and on the preopercle. Also, five rows of
papilla-like spines develop on the body. Larvae of T. maculatus
develop bony ridges on the frontal, parietal and autopterotic
bones. A. lineatus larvae are unique among described soleids in
having an elongate third dorsal ray. Early larvae of .-1. lineatus
are unpigmented but by late preflexion stage have developed
pigment on the head, gut, elongate dorsal ray, dorsal and ventral
body margins and blotches on the dorsal and anal fins. Early
larvae of T. maculatus are heavily pigmented and have three
large blotches in the dorsal finfold and two in the ventral finfold.
In later larvae these blotches become dusky bars that overlie
the nearly solid background pigment.
Soleines have a large head and jaws as in achirines but the
eye is relatively smaller and the dorsal hump is less prominent
(Fig. 356). Also, soleines are less deep-bodied and the gut oc-
cupies a relatively smaller portion of the body mass; in many
soleine species the rounded gut mass protrudes well beyond the
ventral profile. Pigmentation is highly varied ranging from species
of Aseraggodes which lack pigment to species such as Solea
solea. Pegusa lascans, Microchirus variegatus and Euryglossa
pan which are solidly covered with melanophores. A typical
pattern appearing in many described species consists of a series
of melanophores along the dorsum, ventrum and horizontal
septum, and melanophores on the head, gas bladder and fin-
folds (Fig. 356).
Cynoglossidae {Fig. 357). — Two subfamilies, Symphurinae and
Cynoglossinae, are recognized in the family. The first larval
descriptions of the former are of Symphurus lacleus, S. ligulata
and S. pusilla (Kyle, 1913). Hildebrand and Cable (1930) de-
scribed a series as 5. plagiusa. but Olney and Grant (1976)
described a different series as S. plagiusa and pointed out that
Hildebrand and Cable's descriptions must refer to another
species. Pertseva-Ostroumova ( 1 965) ascribed a larval series to
S. orientalis and we have identified eggs and larvae of .S'. atn-
cauda. Larval series or metamorphosing specimens have been
ascribed to at least 1 1 types of cynoglossines; however, most of
these are incomplete series and identifications are tentative (Ta-
ble 178). Most cynoglossids are less than 2.5 mm at hatching;
Fig. 353. Larvae of Pleuronectidae. (A) Glyptocephatus zachinis, 48.7 mm, redrawn from Ahlstrom and Moser, 1975; (B) G. stelleri. 24.6
mm, from Okiyama and Takahashi, 1976; (C) Tanakius kitaharai. 15.8 mm, ibid; (D) Microstomus achne. 8.8 mm, ibid; (E) Embassichthys
balhybius. 16.2 mm, from Richardson, 1981b.
AHLSTROM ET AL.: PLEURONECTIFORMES
665
666
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Fig. 354. Transforming specimens of Pleuronectidae. (A) Hippoglossus stenolepis. 24.0 mm, original; (B) Eopseltajordam. 16.2 mm, CalCOFl
5104, Sta. 70.55; (C) Lyopsetta exilis. 22.0 mm, from Ahlstrom and Moser, 1975; (D) Pleuronichthys ritteri, 10.0 mm, from Sumida et al., 1979.
i
AHLSTROM ET AL.: PLEURONECTIFORMES
667
Fig. 355. Larvae and transforming specimens of Pleuronectidae. (A) Samaris crislatm, 6.4 mm, from Pertseva-Ostroumova, 1965; (B) 5.
crislatus. 8.7 mm, ibid; (C) Samariscus sp., ca. 24.0 mm; (D) Poecilopsetta hawaiiensis. ca. 29.0 mm; (E) Pelolretis llavilalus. 4.3 mm, redrawn
from Rapson, 1940. C and D from P. Struhsakcr. unpublished.
an exception is C. abbreviatus which has a relatively large egg.
Notochord flexion and transformation occur at larger sizes in
symphurines compared with cynoglossines and some Symphii-
rus have an extended larval stage that exceeds 30 mm in length
in having a large deep head and tapering body, but the jaws are
relatively smaller in cynoglossids and the body is more attenuate
(Fig. 357). The gut mass protrudes beyond the ventral profile
and in some species it trails posteriad. In S. laclea a conical
(Table 178). Cynoglossid larvae are similar to those of soleids structure is attached to the trailing gut coil (Kyle, 1913). Cy-
668
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
c
IV^ifi^V^^*^ • -^'/^i^s.
D
Fig. 356. Larvae and transforming specimens of Soleidae. (A) Trinecles maculalus. 2.0 mm. from Hildebrand and Cable, 1938; (B) Achirus
linealus. 3.1 mm. from Houde et al.. 1970; (C) Solea solea. 7.5 mm, from Ehrenbaum, 1905-1909; (D) MUrochirus vanegatus. 10.0 mm, from
Petersen, 1909; (E) Eun-glossa pan. 4.6 mm, from Jones and Menon, 1951; (F) Solea mala. 4.7 mm, from Jones and Pantulu, 1958; (G)
Microchirus ocellatus. 5.1 mm, from Palomera and Rubies, 1977; (H) .iuslroglossus microlepis. 6.6 mm, from O'Toole, 1977; (I) Heleromycteris
japonicus. 4.9 mm. from Minami. 1981b; (J) .Aseraggodes whitakeri. ca. 27.0 mm, from P. Struhsaker, unpublished.
AHLSTROM ET AL.: PLEURONECTIFORMES
669
'^'^aa^
H
~'^^. ■■ 1
o^- •■
Fig. 357. Larvae of Cynoglossidae. (A) Cynoglossus abbreviatus. 5.0 mm. from Fujita and Takita. 1965; (B) C monopus. 7.0 mm. from
Balaknshnan. 1963; (C) C. macrostomus. 4.5 mm, from Seshappa and Bhimachar. 1955; (D) Syinphurus hgulala. 10.5 mm. from Kyle. 1913;
(E) S. alncauda. 4.0 mm. onginal. CalCOFI; (F) S atncauda, 6.5 mm. onginal. CalCOFl; (G) S. alncauda. 12.8 mm. original, CalCOFI; (H) 5.
plagiusa. 6.2 mm, redrawn by Fahay (1983) from Olney and Gram. 1976; (I) 5. lactea. 18.0 mm. from Padoa, 1956k.
noglossid larvae develop a crest consisting of elongate anterior
dorsal rays. 2 rays in Cynoglossus and usually 4 or 5 in Syin-
phurus. Pectoral fins are present during the larval period, but
do not develop rays and disappear at metamorphosis. One
species, S. ligulata. develops elongate third and fourth pelvic
rays (Kyle. 1913; Padoa. 1956k).
Pigmentation in early larvae of Cynoglossus consists of 4-5
opposing blotches along the dorsum and ventnim. pigment on
670
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
the head, gut and gas bladder. In some species, large blotches
in the finfold distal to the dorsal and ventral midline blotches
give the larvae a barred appearance. In later stages the midline
blotches become more numerous and some species develop a
series along the horizontal septum. Early larvae of Symphurus
have small melanophores along the ventral midline, and in some
species, also along the dorsal midline. Most species have a single
bar posteriad on the tail and at least one, 5. athcauda, has large
blotches at the finfold margins. The head (particularly ventrally),
gut, gas bladder and horizontal septum become pigmented and
later-stage larvae have pigment patterns similar to Cynoglossus
species.
Metamorphic stages
Pleuronectiforms undergo a remarkable metamorphosis dur-
ing which one of the eyes, the left in dextral and the right in
sinistral species, migrates around or through the head to a po-
sition dorsal to the non-migrating eye. Metamorphosis occurs
over a wide size range among flatfishes, from about 5 mm in
achirine soles (Houde et al.. 1970) to greater than 120 mm in
some bothines (Amaoka, 1971). Capture of specimens of the
enormous flatfish larva observed by Barham ( 1 966) from a div-
ing saucer may double the maximum size for flatfish larvae.
Most flatfishes metamorphose within the range of 10-25 mm
(see preceding sections and Tables 173-178); the size interval
over which the process occurs is smaller in species which meta-
morphose at a small size.
Metamorphosing specimens are relatively rare in plankton
collections because 1) the process is transitory, 2) avoidance is
increased at larger sizes, and 3) metamorphosing individuals
may change habitat. Existing information indicates a variety of
mechanisms of eye migration among flatfishes. In groups where
the dorsal fin origin in larvae is at the posterior margin of the
eye or more rearward (psettodids, citharids, scophthalmids, most
paralichthyids, pleuronectids), a depression forms in the inter-
ocular region and the eye migrates over the dorsal midline an-
terior to the fin origin. Subsequently the dorsal fin extends for-
ward to its adult position (except in psettodids). In larvae of
bothids and the paralichthyid genera Cyclopsetta. Syaciiun and
Cithanchlhys (some species), the dorsal fin is attached to the
skull anterior to the eye and, during metamorphosis, the eye
migrates through a slit which forms between the fin base and
the skull. In some metamorphosing soleids the dorsal fin projects
forward above the snout and the eye migrates through the space
between this protuberance and the skull; subsequently the fin
projection fuses to the skull (Houde et al., 1970; Palomera and
Rubies, 1977; Minami, 1981b). Seshappa and Bhimachar( 1955)
described the process of eye migration in a captive specimen of
Cynoglossus macrostomus. Just before eye migration a fleshy
hook-shaped protuberance grew forward from the region of the
head anterior to the dorsal fin origin. The right eye migrated
through the space between the protuberance and the skull, after
which the fleshy appendage fused to the dorsal region of the
skull. The entire process took place over a 5-hour period during
the night. A similar structure appears on advanced larvae of an
unidentified cynoglossid illustrated by John (1951b) and this
mechanism of eye migration may be widespread among cyno-
glossids.
During eye migration in flatfishes a number of other meta-
morphic events occur: 1) larval spines are lost, 2) elongate rays
assume their j uvenile proportions, 3) gut protrusions are brought
into the body cavity and internal organs are rearranged, 4) gas
bladder, if present, is lost, 5) pectoral fins develop rays, except
in cynoglossids, some soleids, some bothids and Mancopsetta,
where (one or both) fins are lost altogether during this period,
6) larval pigment patterns are replaced by juvenile patterns, 7)
ossification of the vertebral column and other bony structures
is completed, 8) intermuscular bones appear in bothids. and 9)
scales form.
(H.G.M., B.Y.S.) National Marine Fisheries Service,
Southwest Fisheries Center, P.O. Box 271, La Jolla,
California 92038; (K.A.) Faculty of Fisheries,
Hokkaido University, Hakodate, Japan; (D.A.H.) De-
partment OF Marine Sciences, University of Puerto
Rico, Mayaguez, Puerto Rico.
Pleuronectiformes: Relationships
D. A. Hensley and E. H. Ahlstrom
BASICS of the current working model for evolution of pleu-
ronectiforms were proposed by Regan (1910, 1929) and
Norman (1934). In his monograph, Norman treated the floun-
ders (Psettodidae, Bothidae, Pleuronectidae), and though he did
not publish a revision of the remaining pleuronectiforms, his
key and classification of the soleoids were published posthu-
mously (1966). Norman's model and classification with the
modifications of Hubbs (1945), Amaoka (1969), Futch (1977),
and Hensley (1977) represent the most recent, detailed hypoth-
esis for pleuronectiform evolution. We will refer to this as the
Regan-Norman model (Fig. 358) and classification (preceding
article, this volume) and consider it the working hypothesis to
be reexamined using adult, larval, and egg characters.
Formation of the Regan-Norman model involved an eclectic
approach, i.e., a combination of phyletic and phenetic methods.
Although some of the groups currently recognized appear to be
based on synapomorphies, many are clearly based on symple-
siomorphies and were recognized as such by the authors. This
search for horizontal relationships among pleuronectiforms us-
ing eclectic methods, with one exception, has been the only
approach used in this group. The exception is the recent work
of Lauder and Liem (1983) in which a cladogram for flatfishes
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
671
Psettodidae
Scophlhalmidae
E
Bothidae
Citharidae
V V
Soleidae Cynoglossidae
Pleuronectidae
Pleuronectoidei
Soleoidei
Psettodoidei
Fig. 358.
(1969).
Current hypothesis for interrelationships of pleuronectiform fishes. Based on Norman (1934, 1966), Hubbs (1945), and Amaoka
is presented. However, these authors present this as a tentative
hypothesis and admit that the interrelationships expressed are
still problematic. Most of the character states they use are re-
ductive, few characters were analyzed, and the authors were
understandably unaware of recent character surveys, since much
of this information is unpublished.
We have made the assumption that the order Pleuronecti-
formes is monophyletic and the sister group is the remaining
percomorph fishes (sensu Rosen and Patterson, 1 969 and Rosen,
1973). Although the monophyly and origin of the group is still
open to question and hypotheses of multiple origins have been
proposed (e.g., Kyle, 1921; Chabanaud, 1949; Amaoka, 1969),
a monophyletic model with a percomorph sister group still ap-
pears to be the most parsimonious. In other words, with the
information available, there appears to be no need to hypoth-
esize multiple origins for flatfishes; to do so demands the inclu-
sion of a great deal of convergence.
Relationships
The following discussion of relationships within the pleuro-
nectiforms is cursory and preliminary. In fact, it asks more
questions than it answers and illustrates that more work (par-
ticularly osteological) is needed in certain groups before the
order can be subjected to an in-depth cladistic analysis. Until
this work is completed, it is premature to offer a new hypothesis
of interrelationships for the entire order.
Adult characters
Several criteria were used for selecting characters for discus-
sion: (1) amount of information available on the distribution
of character states; (2) characters commonly used in the past to
define groups of pleuronectiforms; (3) those for which our
knowledge of distributions of states is limited, but appear to
indicate groupings different from those hypothesized in the
working classification and which need additional study; and (4)
characters which are well known in certain groups and are po-
tentially useful for elucidating relationships within these groups.
Characters and character complexes used in this study are dis-
cussed below. Characters and states are presented in Table 179.
Optic chiasma.—The relationship between the optic chiasma
and ocular asymmetry of pleuronectiforms has been investigated
by several workers beginning mainly with the work of Parker
(1903). Hubbs (1945) examined this relationship further and
presented all data from previous studies. Parker found that most
fishes have a dimorphic optic chiasma, i.e., the nerve of the left
672
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
or right eye is dorsal with about equal frequency (state referred
to here as truly dimorphic). Exceptions to this are species of
paralichthyids (sinistral) and pleuronectines (dextral) where the
right or left optic nerve, respectively, is always dorsal, even in
reversed individuals, i.e., the optic chiasma is monomorphic.
The Soleidae and Cynoglossidae, however, retain a truly di-
morphic optic chiasma. Subsequent work by Regan (1910) and
Hubbs (1945) showed that in the indiscriminately dextral or
sinistral Psettodes the optic chiasma is also truly dimorphic. In
addition, Hubbs presented evidence of a third state, at least in
Citharoides (sinistral), where the nerve of the migrating eye is
dorsal even in reversed individuals. He thus interpreted the
Citharidae as having a basically dimorphic optic chiasma and
predicted the same for scophthalmids, although apparently no
one has examined a reversed scopthalmid to test this prediction.
A truly dimorphic optic chiasma as found in Psettodes and the
soleoids has been interpreted as plesiomorphic for pleuronec-
tiforms. The type of optic chiasma found in Citharoides and
predicted for scophthalmids (i.e., nerve of the migrating eye
always dorsal) was interpreted as an intermediate state between
the truly dimorphic and the monomorphic chiasmata as found
in pleuronectoids. We agree with this interpretation of polarity.
However, some plesiomorphic states have been used to define
groups: Psettodidae, truly dimorphic; Citharidae, basically di-
morphic; Scophthalmidae, predicted to be basically dimorphic;
and Soleoidei, truly dimorphic.
Major problems exist with the use of the optic chiasma for
phylogenetic inference. One of these concerns the feasibility of
actually determining which state exists in a group. Demonstrat-
ing the occurrence of truly dimorphic chiasmata is relatively
simple. All that is needed is to show that either optic nerve is
dorsal regardless of which eye has migrated; reversed individuals
are not necessary. To demonstrate occurrence of the basically
dimorphic state, reversals are needed and the nerve of the mi-
grating eye must always be dorsal. Likewise, reversed individ-
uals must be examined to show a monomorphic chiasma. Here
the nerve to the right eye must be dorsal in all individuals
(including reversals) of normally sinistral species and the nerve
of the left eye must be dorsal in all individuals of normally
dextral species. When one actually examines the data for this
character (see Hubbs, 1945), states have been determined for
very few pleuronectiform groups. The occurrence of the basically
dimorphic state in the Citharidae was demonstrated in only one
species. Of greater significance, however, is the fact that a mono-
morphic state has been shown for very few pleuronectoid species.
Within the pleuronectoids it has been widely assumed that all
paralichthyids, bothids, and pleuronectids have monomorphic
optic chiasmata, and that because of this they are monophyletic
and not closely related to the soleoids (truly dimorphic). It is
worthy of note here that a monomorphic optic chiasma has
never been demonstrated for four pleuronectid subfamilies
(Poecilopsettinae, Rhombosoleinae, Samarinae, Paralichthod-
inae), the Bothidae, or the paralichthyid genus Thysanopsetta.
Ocular asymmetry— T\\K character (sinistral, dextral, indis-
criminate) is obviously interrelated with the optic chiasma in
certain groups, i.e., those with basically dimorphic and mono-
morphic chiasmata. The evolution of ocular asymmetry and its
relationship to the optic chiasma is not well understood, al-
though there is one major hypothesis (Norman, 1934; Hubbs,
1945) which states that primitively, pleuronectiforms were in-
discriminate in ocular asymmetry and the optic chiasma was
truly dimorphic. Soleoids became discriminate (soleids dextral
and cynoglossids sinistral), but retained a truly dimorphic chias-
ma. Psettodids remained indiscriminate and truly dimorphic.
Citharids and presumably scophthalmids became discriminate
(scophthalmids and citharines sinistral and brachypleurines
dextral) but retained some ontogenetic plasticity in regard to
the optic chiasma, since reversed individuals still have the nerve
of the migrating eye dorsal (basically dimorphic). The remaining
pleuronectoids became discriminate (Paralichthyidae and Both-
idae sinistral and Plueronectidae dextral) and evolved a mono-
morphic chiasma. The only exceptions with regard to ocular
asymmetry are certain indiscriminate paralichthyids and pleu-
ronectines. However, most of these indiscriminate pleuronec-
toids have been shown to have a monomorphic optic chiasma
(a possible exception is Tephnnectes). It would thus appear that
indiscriminate ocular asymmetry in pleuronectoids developed
secondarily from discriminate ancestors (Hubbs and Hubbs,
1945).
Making phylogenetic interpretations from two states of ocular
asymmetry is difficult or impossible without corroborative evi-
dence. Thus, a statement to the effect that two or more dextral
(or sinistral) pleuronectoid groups are most closely related to
each other because they are dextral (or sinistral) without addi-
tional evidence of synapomorphies is circular, and may lead to
the recognition of polyphyletic groups. This reasoning was the
basis for the proposed close relationship in the Regan-Norman
model between the Pleuronectinae and the remaining pleuro-
nectid subfamilies (Poecilopsettinae, Rhombosoleinae, Samar-
inae, Paralichthodinae) and for treating the genera Mancopsetta
and Thysanopsetta as members of the Bothidae and Paralich-
thyidae, respectively.
Ribs and intermuscular bones. — In pleuronectiforms that pos-
sess ribs, these appear to be homologous with the pleural and
epipleural ribs of other teleosts, and the presence of these bones
should be considered plesiomorphic for the order. Two groups
lack both series of ribs, the Achirinae and apparently the Cyn-
oglossidae. Chabanaud (1940) reports epipleural ribs in some
cynoglossids but mentions no genera or species. We have not
seen them in cleared-and-stained Symphurus species or in ra-
diographs of several Cynoglossus species. Although it is still
commonly believed that all soleoids lack both series of ribs (e.g..
Nelson, 1976; Lauder and Liem, 1983), Chabanaud (1940, 1941)
found short epipleural ribs in Solea, Microchirus, and Aesopta,
and we have seen them in Aseraggodes.
Chabanaud (1940, 1950, 1969) found additional rib-like
bones ("metaxymyostes") in certain pleuronectiforms. Some of
his statements about these were in error, and it is now clear he
was referring to Bothus podas and Samaris cristalus (Hensley,
1977). Amaoka (1969) found these ("intermuscular") bones in
all species of his Bothidae and presented very detailed descrip-
tions of their morphology. One of his primary justifications for
elevating Norman's ( 1934) Bothinae to the family level was the
presence of these bones in the group and their absence in Nor-
man's Paralichthyinae. Norman considered Engyophrys. Tri-
chopsetta, Monolene. Taeniopsetta, and Perissias to be paralich-
thyines. All of these genera have intermuscular bones (Amaoka,
1969; Futch, 1977; Hensley, 1977; pers. observ.) and are con-
sidered here to be bothids.
Bothid intermuscular bones are in five series. Amaoka ( 1 969)
called these series epimerals, epicentrals, hypomerals, and
myorhabdoi (two series). He interpreted three of these (epi-
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
673
merals, epicentrals, hypomerals) as homologous with those of
lower teleosts (see Phillips, 1942). The presence of these bones
was the main reason both Chabanaud (1949a) and Amaoka
(1969) hypothesized that pleuronectiforms were polyphyletic
and that at least the Bothidae, and in the case of Chabanaud
also the Samarinae, were derived from some lower teleostean
group. Hensley (1977) presented arguments for interpreting the
pleuronectiforms as monophyletic and the presence of inter-
muscular bones in at least the Bothidae as being apomorphic.
Chabanaud (1969) described intermuscular bones in Samaris
as being in two series. However, we recently e.\amined a cleared-
and-stained specimen and found differences with Chabanaud's
description. In the abdominal region, rib-like or intermuscular
bones are in three series. Bones of the middle series are un-
branched and in the horizontal skeletogenous septum. Most
bones of the dorsal and ventral series are branched. In the region
of the caudal vertebrae, there are only the dorsal and ventral
series. There are none of the dorsal and ventral myorhabdoi as
found in the Bothidae. Although the three series of bones found
in Samaris resemble the epimerals, epicentrals, and hypomerals
of bothids, a more detailed comparison is required before a
statement about homologies can be made.
Amaoka (1969) interpreted bothids as lacking pleural and
epipleural ribs, but possessing the five series of intermuscular
bones. However, there is another interpretation. It is possible
that Amaoka's epicentrals (limited to the horizontal skeletog-
enous septum of the abdominal region) and abdominal hypom-
erals are homologous to epipleural and pleural ribs, respectively,
of other pleuronectiforms, and that the presence of myorhabdoi,
epimerals, and caudal hypomerals are apomorphic states.
Postcleithra. — The absence of postcleithra was a character state,
apparently apomorphic, used by Norman ( 1 934) and subsequent
authors to distinguish the Soleoidei from the Psettodoidei and
Pleuronectoidei. However, an adequate survey of this character
has never been made among the pleuronectoids. In a preliminary
survey, we found postcleithra absent in certain pleuronectoids,
i.e., the Samarinae and the bothid genera Mancopsetta and Pel-
ecantchthys. Postcleithra are definitely present in the rhombo-
soleines Oncopterus, Azygopus. Ammotretis. and Colistium. but
they may be absent in Pelotretis. Pellorhamphus. and Rhom-
bosolea (Norman, 1934: fig. 25c; Chabanaud, 1949). Although
lack of postcleithra in pleuronectiforms is reductive, their ab-
sence in certain pleuronectoids may indicate a closer relation-
ship between some of these groups and soleoids than hypoth-
esized in the Regan-Norman model. The occurrence of this
specialization in Pelecanichthys is almost certainly an indepen-
dent reduction, since this genus shows several synapomorphies
with other bothids.
Vomerine teeth. — Huhhs (1945, 1946) interpreted the presence
of vomerine teeth as a primitive state for the order, and we
concur. However, Hubbs presented this interpretation as evi-
dence that citharids and scophthalmids were closely related and
represented an intermediate grade in pleuronectoid evolution.
The presence of vomerine teeth cannot be used to infer phy-
logenetic relationships among pleuronectiforms.
Fin spmes— Huhhs ( 1 945, 1 946) presented the distributions for
dorsal, anal, and ventral-fin spines in pleuronectiforms. Psei-
todes is the only genus with dorsal and anal spines. This genus
and the Citharidae are the only flatfishes with ventral-fin spines.
Hubbs properly interpreted their presence in these groups as
plesiomorphic for the order. However, again, he used a hori-
zontal or eclectic approach and inferred a close relationship
between the citharid genera and interpreted the group as an
intermediate grade in pleuronectoid evolution. The presence of
these spines does not indicate phylogenetic (vertical) relation-
ships.
Supramaxillae. — SxxpvdiVmxiWac occur in Psettodes and the cith-
arids Eucitharus and Citharoides (Hubbs, 1945). In Psettodes,
the bones are well developed and apparently present on both
sides. The two citharid genera have them reduced in size, con-
fined to the blind side, or sometimes missing. The presence of
these bones is plesiomorphic for the order and should not be
used to infer phylogenetic relationships.
Ventral-fin placements and base lengths. — Evolution of ventral-
fin asymmetry in pleuronectiforms is not well understood. Most
of our knowledge concerning the relationship between ocular
and ventral-fin asymmetry has come from some rare examples
of reversals in forms with asymmetrical ventral-fin morphology
(see Norman, 1 934). For comparative purposes, i.e., attempting
to determine homologous states, it would appear to be more
correct to compare ocular and blind-side ventral fins between
groups rather than those of the right and left sides (see Hubbs
and Hubbs, 1945). At present, there are several problems in
using ventral-fin morphology to elucidate phylogenetic rela-
tionships. Most work here has dealt only with external mor-
phology and much of this has not been sufficiently detailed or
accurate. What is needed are thorough comparisons of basip-
terygia as well as fins. Due to the paucity of accurate and detailed
studies of these structures in flatfishes, it is not possible to ad-
equately define character states for an in-depth comparison
throughout the order. Thus, ventral-fin characters were not in-
cluded in Table 1 79. What follows is a discussion of general
patterns of ventral-fin morphology.
Ventral fins with short bases and symmetrical placements
have been correctly considered plesiomorphic states in pleu-
ronectiforms, and any type of asymmetry in placement, size,
shape, o;^ meristics as having been derived from symmetrical
states (e.gir Norman, 1934; Hubbs, 1945; Amaoka, 1969). Most
ventral-fin characters used have involved positions of the fins
relative to the midventral line and relative lengths of the fin
bases. Unfortunately, symmetry (plesiomorphic states) in both
of these characters has been used to define groups. Short-based
fins and symmetry or near symmetry in placement and base
lengths occur in Psettodes. the Paralichthyidae (except the Cy-
clopsetta group), the Citharidae, most soleines, most or all Pleu-
ronectinae, and the Poecilopsettinae. States where the ocular
ventral fin is on the midventral line and has a base extending
farther anteriorly than that of the blind side form a continuum.
Thus, groups with the base of the ocular ventral fin only slightly
extended anterior to that of the blind side (origin of blind fin
at transverse level of about the second or third ray of the ocular
fin) are the Samarinae, possibly some Soleinae, Paralichthodes,
the Taeniopsettinae, and Monolenc, groups where the origin of
the ocular fin is farther anterior relative to that of the blind fin
are the Rhombosoleinae, all Bothinae (except Monolene). and
possibly some Soleinae. Two groups, the Scophthalmidae and
Achirinae, have both ventral-fin bases close to or virtually on
the midventral line and the anterior basipterygial processes ex-
tended. The Cyciopsetta group has the ocular fin on the mid-
674
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Table 1 79. Characters and States for Pleuronecttform Groups. Where appropriate states are indicated by underlined letters. See text and
Figs. 359-364 for hypural fusion patterns.
Optic
Ocular
chjasma
asym-
Post-
(Truly
metry
cleithra
Dimorphic,
(Dc.xtral.
Ribs
Intermuscular Bones
(Present.
Basically
Sinis-
(Present, Absent)
(Present
Absent)
Absent
Vomerine
Ventral-
tin
Supra-
Dimorphic,
tral.
Present
teeth
forniul-i
Mono-
Indis-
Epi-
Epi-
Hypo-
Epi-
Myo-
only in
(Present.
(ocular SI
dc/
(Present,
Taxon
morphic)
cnminate)
pleural
Pleural
centrals
merals
merals
rhabdoi
Larvae)
Absent)
blind side)
Absent)
Psettodidae
TD
I
p
P
A
A
A
A
P
P
1, 5/1,
5
P
Citharidae
Brachypleurinae
Brachypleura
9
D
p
P
A
A
A
A
P
P
I, 5/1,
5
A
Lcpidohlepharon
9
D
p
P
A
A
A
A
P
P
I, 5/1.
5
9
Citharinae
Citharoides
BD
S
p
P
A
A
A
A
P
A
1, 5/1,
5
P
Eucitharus
9
S
p
P
A
A
A
A
P
P
I, 5/1,
5
P
Scophthalmidae
9
S
p
P
A
A
A
A
P
P. A
6/6
A
Paralichthyidae
*Cyclopsel!a group
9
S
p
P
A
A
A
A
P
A
5-6/6
A
"Pseudorhomhus group
9
S
p
P
A
A
A
A
P
A
6/6
A
***Paratichthys group
M
S.I
p
P
A
A
A
A
P
A
6/6
A
Tephrinectes
M?
I
p
P
A
A
A
A
P
A
6/6
A
Thysanopsetta
9
S
p
P
A
A
A
A
P
A
6/6
A
Bothidae
Taeniopsettinae
9
S
p?
P?
P
P
P
P
P
A
6/6
A
Bothinae
9
S
p?
P?
P
P
P
P
P. A
A
6/6
A
****MancopseUa
9
s
p
P
A
A
A
A
A
A
7/5-
-7
A
Pleuronectidae
Pleuronectinae
M
D, I
p
P
A
A
A
A
P
A
4-7/4-
-7
A
Poecilopsettinae
?
D
p
P
A
A
A
A
P
A
6/6
A
Paralichthodinae
9
D
9
9
9
9
9
9
9
A
6/6
A
Samarinae
9
D
P?
P?
P
P
P
A
A
A
5/5
A
Rhombosoleinae
9
D
P
P
A
A
A
A
P, A?
A
6-13/0-
-6
A
Soleidae
Soleinae
TD
D
P
A
A
A
A
A
A
A
5/5
A
Achirinac
TD
D
A
A
A
A
A
A
PL
A
3-5/2-
-4
A
Cynoglossidae
TD
S
A
A
A
A
A
A
A
A
0-2/4
A
* Cilhahchrhys. Cychpsella. Etropus. Syacmm
•* Cephalopseira. Pseitdorhombw,. Tarphops.
*** Ancylopsetta, Gastropsetta. Htppvglossma. Lioglossina. Parahchlhys. X'erecundttm. Xystreuiy^
**** Achiropsetia and Neoachiropsella are considered synonyms.
ventral line, but the basipterygium of the blind fin is placed in
a more anterior position than that of the ocular side. Another
unique state is the loss of the blind ventral fin in some genera
of the Rhombosoleinae, although the basipterygium of the blind
side is probably still present. The Cynoglossidae are the only
pleuronectiforms in which the blind ventral-fin base is oriented
along the midventral line and the ocular fin is in a more dorsal
position or absent. In cynoglossids missing the ocular fin, at
least the dorsal process of the left basipterygium is still present.
Vertebral transverse apophyses. — Regan (1910) used the pres-
ence of transverse apophyses on caudal vertebrae as a state to
distinguish his bothid subfamilies Platophrinae and Bothinae
from the Paralichthyinae (=Paralichthyidae with modifica-
tions). Norman ( 1 934) combined the Platophrinae and Bothinae
into his Bothinae and Scophthalminae and again used transverse
apophyses on caudal vertebrae to distinguish the bothines and
scophthalmines from the paralichthyines. Amaoka (1969) used
the presence of these structures to define his Bothidae and dis-
tinguish them from the other sinistral flounders he treated (Para-
lichthyidae, Citharidae, Psettodidae). Hensley (1977) and Futch
(1977) found transverse apophyses in Engyophrys, Trichopsetta,
and Monolene and suggested this as a character state indicating
these genera were more closely related to the Bothidae than the
Paralichthyidae. We have since found them in Perissias. As
previously slated, Norman (1934) had placed these four genera
in the Paralichthyinae. Amaoka (1969) presented the most de-
tailed descriptions of these structures. Basically, there are two
pairs of transverse apophyses on the vertebrae, an anterior and
a posterior pair. They are found on many abdominal and most
caudal vertebrae. Subsequently, we have found that the trans-
verse apophyses seen by Regan (1910) and Kyle (1921: fig. 32)
in the scophthalmids are very similar to those present in the
Bothidae. They are similar in shape and occur in two pairs.
Amaoka (1969) interpreted the occurrence of these structures
in the Bothidae as indicative of a relationship to some fish group
other than the Percomorpha and used this as evidence that the
Bothidae arose independently from the remaining pleuronec-
tiforms. To support this, he cites the occurrence of similar struc-
tures in anguilliforms (Trewavas, 1932; Asano, 1962).
Recently in a preliminary survey of this character or complex
in other flatfishes, we found transverse apophyses on some ver-
tebrae in the Samarinae, Cynoglossidae, and Soleinae. However,
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
Table 179. Extended.
675
Position
Infra-
of unnary
Position
Articu-
orbital
Firsl
papilla
of vent
Haemal
lation of
lateral-
neural
(Midven-
(Midven-
arch on
parhypural
Haemal
line
Vertebral
transverse apophyses
spine
tral.
~tral.
parhypural
with termi-
spine
canal on
(Presenl.
Absenl.
Ocular
side.
Ocular
side.
(Presenl,
Absent.
nal half
centrum
Hypural
fusion
Number
on PU2
(Autog-
Branched
ocular
side
(Present.
Absent)
. Absent or
Blind
Blind
Rudimen-
(Present.
pattern
of autogenous
enous,
caudal-lin
(Present.
Antenor
Poslcnor
Reduced)
side)
side)
tary)
Absent)
(1-6)
epurals
Fused)
Total caudal-fin rays
rays
Absent)
A
A
P
M
M
P
P
1
1
A
24-25
15
P
A
A
P
O
O
R?
A
6
1
F
21
13
A
A
A
P
O
O
P?
P
1
2.3?
A
23
15
P
A
A
P
O
O
R?
A
2
2,3?
A
22-
-23 (usually 23)
15
P
A
A
P
O
O
A
A
4
I
F
21
14-15
P
P
P
P
o
M, B
A
A
6
1,2?
F
16-17
13-15
P
P
P
A
B
B
A
A
6
0
F
17
11
A
A
A
P
O
B
R?
A
6
0
F
17-
-18 (usually 17)
10-13
P
A
A
P
O
M. B
A
A
6
0. 1,2?
F
18
13
P
7
7
P
O
M
R''
A
1
2
F
20
14
P?
7
7
P
O
B
R?
A
1
1
F
15-16
0
A?
P
P
A
O
B
A
A
6
0
F
16-
-18 (usually 17)
10-13
A
P
P
A
O
B
A
A
6
0
F
15-
-18 (usually 17)
9-13
A
A
A
AR
O
M
R?
A
1
1
F
14-
-19 (usually 16-18)
13-14
A
A
A
P
O
M, B
A
A
6
1.2?
F
17-24
10-16
P
A?
A?
AR
o
M
R
A
1
1
F
20
14-15
7
7
7
P
9
B
■7
9
1
0
9
16
12
7
P
A
A
o
M
A
A
5
1?
F
16
0-12
A
A
A
AR
o
M. B
A
A
1.4
1
F
17-20
8-15
A
P
A
A
o
M
A
A
4
1
F
16-
-20 (usually 18)
0-18
7
A
A
AR
o
B
A
A
1.2?, 3
1,2?
A
15-
-18 (usually 16)
11-17
P
P
A
AR
M
B
A
A
4
0. 1
F
8-14
0
A
in these groups, they occur only as one pair on the anterior end
of the vertebra. In addition, the Cyclopsetta group has two pairs
of very small lateral protuberances on most vertebrae. How to
interpret the presence of vertebral transverse apophyses in pleu-
ronectiforms is still open to question.
First neural spine.— Amaoka {1969) found that the neural spine
of the first vertebra is missing in the Bothidae and interpreted
this as a synapomorphy for the group, since absence of this spine
is apparently rare or unknown in other teleosts. We have made
a preliminary survey for this in other pieuronectiforms not treat-
ed by Amaoka. Some of this survey was based on radiographs,
and due to the close proximity of the first vertebra and neu-
rocranium, in some groups we are not sure if the first neural
spine is present, absent, or greatly reduced. The states in other
groups are more certain, since some cleared-and-stained ma-
terial was available. A greatly reduced or missing first neural
spine is not limited to the Bothidae (Table 1 79).
Position of the urinary papilla. — AW flatfishes have a papilla on
the posteroventral area of the abdomen near the anal-fin origin.
Schmidt (1915, cited by Norman, 1 934) commented on its po-
sition in flatfishes, claiming it was located on the ocular side in
all species. However, Chabanaud (1934), Hubbs (1945), and
Hubbs and Hubbs (1945) found it to be on the midventral line
in Psettodes. In addition, Hubbs (1945) and Hubbs and Hubbs
(1945) found the papilla on the blind side in the paralichthyid
genera Syacium, Citharichthys. and Etropus. We have found it
in the same position in Cyclopsetta. Another exception here
may be certain cynoglossids. Menon (1977: fig. 45) shows the
urinary papilla on the blind side in a species of Cynoglossus,
but claims it is attached to the first anal-fin ray in all species of
the family. A midventral position for the papilla is generalized
for teleosts and plesiomorphic for pieuronectiforms.
Position o/" ve«?. — Position of the anus in flatfishes has been
reviewed by Norman (1934), Hubbs (1945), and Hubbs and
Hubbs (1945). A midventral position is plesiomorphic for the
order. In flatfishes where the vent is on or near the midventral
line, it is often very difficult to determine what state is repre-
sented. It is on the blind side in several groups, but apparently
on the ocular side only in the Citharidae. Hubbs (1945) inter-
preted the distribution of these states as indicating that deflec-
tion of the vent to the blind side has occurred several times
within the order.
Caudal-fin complex.— The caudal fin and skeleton of many
species of pieuronectiforms have been illustrated and discussed
(e.g., Monod, 1968; Amaoka, 1969). The caudal skeleton of
Psettodes is reported to be the most primitive among living
676
ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM
NS(EP?) UN
PHYP
Fig. 359. Caudal skeleton oi Psetlodes bennctti. Hypural pattern 1.
EP = epural, HY 1-5 = hypurals 1-5, NS = neural spine, PHY = par-
hypural. PHYP = parhypurapophysis, PL) 2, 3 = prcural centrum 2, 3,
THC = terminal half centrum. UN = uroneural. Redrawn from Monod
(1968).
flatfishes. It can be characteinzed as follows (Fig. 359): a par-
hypural with a haemal arch and parhypurapophysis; five au-
togenous hypurals; two pairs of uroneurals, i.e., pairs of stegurals
and splinter bones; two epurals, the first between the neural-
arch remnants of the second preural centrum; terminal half
centrum, i.e., fusion of two ural centra and the first preural
centrum; haemal spine of the second preural centrum autoge-
nous; haemal spine of the third preural centrum fused; and 24-
25 caudal rays, 17 principal, 15 branched. The caudal skeleton
of Psetlodes has been labelled as basically percoid (e.g., Wu,
1932; Monod, 1968; Amaoka, 1969). It should be noted here
that the neural spine of the second preural centrum is interpreted
as probably a captured epural, and that apparently only one free
epural remains. This is one of the more important differences
between Psetlodes and all other pleuronectiforms, which have
a neural spine on the second preural centrum and apparently a
basal number of two epurals. There are at least two hypotheses
which may explain this difference: (1) The earliest pleuronec-
tiforms may have had three free epurals, the anteriormost be-
coming wedged in the neural-arch remnant on the second preur-
al centrum (i.e., captured) and, thus forming a secondary neural
spine. In Psetlodes the remainmg epurals were fused (Amaoka,
1 969) or one was lost, while both were retained in the remaining
flatfishes, at least primitively. (2) The earliest pleuronectiforms
had two epurals, the anteriormost being captured in Psetlodes,
leaving one free epural. In the remaining flatfishes a neural spine
on the second preural centrum was acquired by fusion of this
vertebra with an anterior one bearing a spine. Rosen (1973) has
discussed the second hypothesis to account for secondaiy ac-
quisition of a neural spine on the second preural centrum and
offered as evidence the frequent occurrence of double spines on
the second preural centrum. Such anomalies are frequent in
pleuronectiforms (see Cole and Johnstone. 1902; Barrington,
1937;Chabanaud, 1937;Amaoka, 1969; Okiyama, 1974;Futch,
1977; Fig. 360H). However, although a detailed survey for these
doubled spines has never been done, it appears that doubled
neural spines on this vertebra are just as frequent as doubled
haemal spines.
In spite of the work that has been done on pleuronectiform
caudal osteology, there is still little agreement on interpretation
of some structures. We cannot solve these problems here or
discuss them in great detail. Most of these differences in inter-
pretation concern certain epaxial elements. More detailed com-
parative work needs to be done on these elements before ho-
mologies can be determined. For example, there is one
interpretation that uroneurals occur only in Psetlodes (Ahl-
strom). However, what appear to be remnants of a stegural may
remain in Cilharoides, Lepidoblepharon, Scophlhalmus. and
some achirines (Fig. 361; Amaoka, 1969; Hensley, pers. ob-
serv.). Although sufficient comparative work has not been done
to treat these dorsal structures across all lines of ffatfishes, within
certain groups we can be fairly sure of homologies, due to certain
consistent patterns of placement and shape and to some larval
work where fusions have been observed.
In regard to neural and haemal spines of the second preural
centrum, the parhypural, and hypurals, our knowledge rests on
firmer ground. Characteristics of these structures have been
widely surveyed and there is much more agreement on inter-
pretation of homologous states. We interpret autogenous neural
and haemal spines on the second preural centrum, retention of
a parhypurapophysis and haemal arch on the parhypural. and
articulation of the parhypural with the terminal half centrum
as plesiomorphic for the order.
Several patterns of fusions occur in regard to hypurals 1-4.
Hypural 5 moves to an epaxial position during ontogeny in
flatfishes (Figs. 360, 362), and its fate is more properly discussed
in reference to fusion (or lack of it) with epurals. The most
primitive condition is where hypurals 1-4 are not fused to the
terminal half centrum or among themselves (pattern 1; Figs.
359, 363 upper).
There are three patterns which are slightly different from each
other. The interpretation of these is not so obvious, and we are
hesitant here to make statements concerning homologies be-
tween groups. One of these (pattern 2) is where hypurals 3 and
4 are fused to the terminal half centrum (Fig. 36 1 ). This pattern
is shown by Citharoides and apparently some Achirinae. In
some achirines, a somewhat different pattern (3) occurs where
hypurals 2, 3, and 4 are fused to the terminal half centrum (Fig.
363 middle). A fusion of hypurals 1-4 to the terminal half
centrum (pattern 4) is found in the Soleinae, Cynoglossidae, one
cithand (Eucilharus), and two genera of Rhombosoleinae (Pel-
torhamphus. Rhoinbosolea\ Figs. 362, 363 lower). Caudal-fin
development in a soleine is illustrated in Fig. 362.
Another pattern of hypurals (5) is unique to the Samarinae
(Fig. 364). There are two ways to interpret this pattern. Here
the central hypurals (2 and 3 or 2-4) are fused to the terminal
half centrum. However, unlike the patterns previously de-
Fig. 360. Caudal-fin structure of Engyophrys senla larvae (A-F). juveniles and adults (G-H). Standard lengths of specimens: (A) 4.6 mm; (B)
5.5 mm; (C) 7.0 mm; (D) 7.6 mm; (E) 7.7 mm; (F) 15.3 mm; (G) 45.7 mm; (H) 82.4 mm. NC = notochord. other abbreviations as in Fig. 359.
Redrawn from Hensley (1977).
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
677
EP*HY5
HY3*4
0 5 mm
HY3 + 4
HY1*2
S HY3'4
= HY1*2
PHY
0.5mm
EP HY5
05 mm
678
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
THC •HY3t4
Fig. 36 1 . Caudal skeleton of Citharoides macrolepis. Hypural pat-
tern 2. HAR = haemal-arch remnant, other abbreviations as in Fig.
359. "V" on distal end of fin ray indicates dorsal- and ventralmost
branched ray.
scribed, in the samarines hypural 1 does not articulate with the
terminal half centrum.
The last pattern of hypurals (6) is characterized as follows
(Figs. 360, 364 middle and lower); hypurals 1 and 2 are fused
together forming one element which articulates with the pos-
teroventral surface of the terminal half centrum; and hypurals
3 and 4 are fused together and to the terminal half centrum.
This pattern occurs in the Pleuronectinae, Paralichthyidae (ex-
cept Tephrinectes and Thysanopsetta), Scophthalmidae, one
citharid (Brachypleura), and the Bothidae (except Mancopsella).
We interpret this pattern as homologous between these groups,
derived, and indicative of a monophyletic origin. We will refer
to these fishes as the bothoid group. Caudal-fin development in
a bothid is illustrated in Fig. 360.
Although there is still some doubt concerning interpretations
of certain epaxial caudal elements in flatfishes, some patterns
are apparent. Most of the information indicates that at least in
most pleuronectiform groups, the basal epural number is two.
However, there is a small third element that appears in many
species (Fig. 36 1 ; first uroneural of Amaoka, 1 969). This element
does not appear to be paired and its interpretation and fate in
some groups is questionable. The two larger epural elements are
still present in some flatfishes (Figs. 361, 363 upper), the cith-
arids Lepidoblepharon and Citharoides and the paralichthyid
Tephrinectes. The fate of these from the perspective of the entire
order is questionable. However, it is obvious that these epurals
have been reduced to one or zero in several groups. Which of
these reductions are homologous is unknown. Within groups
defined by other specializations, however, we are probably jus-
tified in assuming these epural reductions took the same course
and are homologous states.
Although space does not allow a more detailed discussion of
other caudal-fin characters, some obvious trends should be men-
tioned; Symmetrization— There is a marked trend among flat-
fishes toward dorsoventral symmetry in the caudal fin and skel-
eton. This has occurred by various types of fusions, losses, and
secondary divisions of elements. These secondary divisions oc-
cur as scissures of varying depths in many caudal elements (Figs.
360H, 362F, 363 lower, 364 upper). Reduction of total and
branched caudal rays— It has long been recognized that more
primitive flatfishes tend to have larger numbers of total and
branched caudal rays. Thus, Psettodes has a total caudal ray
count of 24-25, 15 of which are branched. In many groups,
caudal rays have been reduced to less than 18 and branched
rays to 0-13.
Infraorbital lateral-line canal on ocular side. — In his study of
sinistral flounders (i.e., Psettodidae and Pleuronectoidei) of Ja-
pan, Amaoka (1969) found ocular infraorbital bones present in
the Psettodidae, two citharid genera (Citharoides. Lepidobleph-
aron), and the Paralichthyidae; they were absent from Japanese
bothids. We have since done some survey work on this character
in other groups not treated by Amaoka and found ocular in-
fraorbital bones missing in additional groups (Table 179).
Examination of the Regan-Norman model
using adult characters
In the following discussion, the groups and classification re-
sulting from the current model for pleuronectiform evolution
will be reexamined. The limited analysis presented here sheds
much doubt on the monophyly of many of the currently rec-
ognized groups and their interrelationships. In a few cases, the
evidence favoring different interpretations is so strong that these
should be recognized in classifications. However, most of this
analysis has produced questions and alternative suggestions that
need additional study.
Psettodoidei, Psettodidae. —Nearly all of the character states used
to define this group (Psettodes. two species) are symplesio-
morphies or have been interpreted as such. Two exceptions, gill
arches with groups of teeth and barbed jaw teeth, are states that
Hubbs (1945) proposed as synapomorphies. Although we have
no reason to doubt that Psettodes is a natural group, it should
be redefined using character states which have been shown to
be synapomorphies.
Soleoidei.— The diflferences between the Soleoidei and Pleuro-
nectoidei were noted and expressed in important classifications
before the works of Regan and Norman (e.g., Jordan and Ev-
ermann, 1896-1900) and they are obviously evident in the cur-
rent model and classification. In most previous systematic re-
search on pleuronectiforms, the author has concerned himself
with one or the other group and assumed that the two were
related only through a common ancestor near the early pleu-
ronectiform line. The possibility, for example, that some so-
leoids may be most closely related to some pleuronectoids has
only rarely been addressed. In any cladistic analysis of pleuro-
nectiform interrelationships, character states used to unite the
soleoids will need to be reinterpreted. Some character states
Fig. 362. Caudal-fin structure of Solea solea larvae (A-C), juveniles and adults (D-F). Total lengths of specimens: (A) 6.0 mm; (B) 6.8 mm;
(C) 8.1 mm; (D) 1 1.5 mm; (E) 18 mm; (F) 470 mm. HA = haemal arch, NA = neural arch, other abbreviations as in Figs. 359, 360. Redrawn
from Fabre-Domergue and Bietrix (1905).
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
679
PUS-
PUS
HY5
PHY
0.5 mm
0.5 mm
HY1 HY4
HY3
HY4
-HY3
B
0.5 mm
0.5 mm
HY4
HYS/ Hys
NS EP
PHY
HY5
THC+HY 1-4
0.5 mm
H
0.5 mm
680
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
HY5
_THC + HY2-4
HY 5
THC + HY 1
Fig. 363. Caudal skeleton of Tephnnectes sinensis. Hypural pattern
1 (upper); caudal skeleton of Tnnectes fimbhata. Hypural pattern 3
(middle); and caudal skeleton of Rhombosolea pleheia. Hypural pattern
4 (lower). Abbreviations as in Fig. 359. "V" on distal end of fin ray
indicates dorsal- and ventralmost branched ray.
used as evidence that soleids and cynoglossids are most closely
related are plesiomorphic for the order (symmetrical nasal or-
gans, dimorphic optic chiasmata), found in some pleuronectoids
but dismissed as parallelisms [lower jaw not prominent, absence
of postcleithra, several "soleoid characters" found in rhombo-
soleines, (see Norman, 1934)], or are incorrect (absence of all
ribs). Other states used to unite the soleoid families include; (1)
a preopercular margin covered by skin and scales; and (2) skin
covering the dentary and interopercular bones being continuous
across the chin, hiding the isthmus and branchioslegal rays (Nor-
man, 1966). A covered preopercular margin is not limited to
soleoids; it occurs in some rhombosoleine genera (Chabanaud,
1949; Hensley, pers. observ.). The second state as well as the
absence of pleural ribs are possible synapomorphies for the
group.
Cynoglossidae.—lhtrt is little doubt that the tonguesoles are
monophyletic. They are unique in having the ventral fin of the
blind side oriented along the midventral line and the ocular fin
placed more dorsally or missing. The relationship of this family
to other groups, however, is obscure (see Soleidae).
Soleidae.—The main character state proposed as uniting the
two soleid subfamilies (Soleinae, Achirinae) appears to be that
all species are dextral. This is still a poorly known group, and
we are not prepared to make much of a contribution here. How-
ever, there are some marked differences between these subfam-
ilies. In several characters, the Achirinae are more primitive
than originally thought. Some species have hypural pattern 1 ,
the most primitive. In species where hypural fusions have oc-
curred, the first hypural remains free and articulates with the
terminal half centrum (Fig. 363 middle). The haemal spine of
the second preural centrum is autogenous (i.e., the plesiomorph-
ic state for the order) in achirines. Uroneurals may still be pres-
ent in some species. Although postcleithra are lacking in adult
soleoids, at least one achirine species has them during larval
development (Futch et al., 1972). Soleines differ from achirines
in these characters in that they show what appear to be more
derived states. The Soleinae have hypurals 1-4 fused to the
terminal half centrum (Fig. 362F), the haemal spine of the sec-
ond preural centrum is attached, there is no indication of uro-
neurals, and postcleithra have not been reported in larvae or
adults. Soleines share these states with the Cynoglossidae. In
addition, both groups have vertebral transverse apophyses, which
are missing in achirines. The possibilities that the Soleidae are
not monophyletic and the Soleinae are more closely related to
the Cynoglossidae should be more thoroughly explored.
Pleuronectoidei. — Some of the character states used to define
this group are plesiomorphic for the order: (1) preoperculum
with free margin; (2) presence of postcleithra; and (3) presence
of pleural and epipleural ribs. Some apomorphic states for the
order are not limited to pleuronectoids; e.g., loss of dorsal and
anal spines. The Regan-Norman model has used the position
of the nasal organ of the blind side to separate pleuronectoids
from soleoids and psettodids. In pleuronectoids, this nasal organ
follows the migrating eye during metamorphosis. After meta-
morphosis, it remains near the dorsal edge of the head. This
was interpreted as a specialization of pleuronectoids, except that
this state does not occur in all Rhombosoleinae (i.e., nasal organs
remain symmetrically placed). Thus, it is not a synapomorphy
for the group, unless it can be shown that the nasal-organ sym-
metry in these rhombosoleines was secondarily derived from
the asymmetrical state. We have not done a survey of nasal-
organ symmetry, but incidental observations indicate that the
supposed differences between these states (i.e., symmetrical vs
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
681
asymmetrical placement) are not as great as formerly thought.
Loss of a truly dimorphic optic chiasma would appear to be the
only synapomorphy proposed to date uniting the pleuronec-
toids. However, as previously discussed, a basically dimorphic
or monomorphic optic chiasma has been demonstrated in very
few pleuronectoid species.
One might expect that we are well informed about the inter-
relationships among pleuronectoids. Unfortunately, all of the
past work has used the eclectic approach. Thus, scophthalmids
and citharids have been related horizontally as primitive pleu-
ronectoids, and bothids, paralichthyids, and pleuronectids as
higher groups. Again, an important character here is the optic
chiasma. Two states were recognized in pleuronectoids: (1) the
primitive one (for pleuronectoids) where the nerve of the mi-
grating eye is always dorsal regardless of eye position (i.e.. the
basically dimorphic state); and (2) the monomorphic state char-
acteristic of "'higher" pleuronectoids where the chiasma is fixed
regardless of reversals. It has been assumed that all bothids,
pleuronectids. and paralichthyids show the monomorphic state.
Some evidence from other characters indicates this assumption
is not valid.
Due mainly to the work of Amaoka (1969) and one of us
(Ahlstrom), we have a good survey of the caudal-fin complex
of pleuronectoids. Patterns of hypurals 1^ are fairly well known.
The distributions of these patterns call into question much of
the current evolutionary model and classification of the Pleu-
ronectoidei. There are five patterns of hypurals in this group as
defined in the Regan-Norman model: Pattern 1 (Fig. 363 up-
per)—This is plesiomorphic for the order. Pattern 5 (Fig. 364
upper)— This pattern is limited to the Samarinae. We interpret
this pattern as a synapomorphy uniting the samarines. Pattern
2 (Fig. 361)— Within pleuronectoids this pattern seems to be
limited to Citharoides. It is apparently derived from pattern 1.
Pattern 6 (Figs. 360, 364 middle and lower)— This is an apo-
morphic pattern which is very distinctive. We consider it ho-
mologous in pleuronectoids where it occurs and a synapomor-
phy uniting these groups. Again, we are calling this group the
bothoids and it includes the Pleuronectinae. Paralichthyidae
(except Tephrinectes and Thysanopsetta). Scophthalmidae,
Bothidae (except Mancopsetta), and Brachypleura. Pattern 4
(Fig. 363 lower)- Within the pleuronectoids this pattern is lim-
ited to certain genera of Rhombosoleinae and Eucitharus. Based
on other characters, the homology of pattern 4 between these
groups is probably not true.
Citharidae. — Many character states used to define this family
(Hubbs, 1945. 1946) are plesiomorphic for the order: (1) reten-
tion of pelvic spines; (2) retention of supramaxillae (Eucitharus
and Citharoides): (3) urinary papilla close to anus; (4) no union
of branchiostegals; (5) retention of vomerine teeth {Eucitharus,
Brachypleura. Lepidoblepharon); and (6) retention of short-based
ventral fins. Some are plesiomorphic for the Pleuronectoidei:
( 1 ) basically dimorphic optic chiasma (at least in Citharoides);
(2) gill membranes showing some degree of union, but still fairly
widely separated; and (3) loss of dorsal and anal-fin spines. The
only possible character state proposed to date that could be
interpreted as a synapomorphy for this family is the position of
the anus on the ocular side. Although we have not examined
many specimens for this character, it appears that deflection of
the anus to the ocular side is probably slight. Amaoka (1972b)
examined Brachypleura and attempted to redefine the Cithar-
idae. However, he still showed no synapomorphies for the group.
EP(HY5''l
HY5(4''I
THC * HY2 -4(2+371
HY5 * EP
THC*HY3 + 4
HYl +2
Fig. 364. Caudal skeleton of Samanscus iriocellatus. Hypural pat-
tern 5 (upper), caudal skeleton of Cilhanchthys macrops. Hypural pat-
tern 6 (middle), and caudal skeleton of Hippoglossina oblonga. Hypural
pattern 6 (lower). SR = splinter ray. other abbreviations as in Fig. 359.
"V on distal end of fin ray indicates dorsal- and ventralmost branched
ray.
682
ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM
The family Citharidae as presently defined is a grade. Ex-
amination of the caudal osteology has shown two derived and
one plesiomorphic pattern of hypurals. Lepidoblepharon shows
pattern 1, which is plesiomorphic for the order. Citharoides
shows pattern 2, a derived pattern (Fig. 361). This pattern could
represent a state on a line leading toward pattern 6, which is
shown by Brachypleura. Eucitharus shows pattern 4, which pos-
sibly developed independently in some rhombosoleines. The
most obvious result of this is that Brachypleura belongs to the
bothoid group, which shares the derived hypural pattern 6. In
this interpretation, the character states shown by Brachypleura
that are primitive for the order (e.g., vomerine teeth, ventral-
fin spines) are also primitive for bothoids.
Scophthalmidae.— Based on ventral-fin morphology, the Scoph-
thalmidae appear to be monophyletic. There are certain simi-
larities in ventral-fin morphology between this family and the
achirines, but these are probably superficial. Scophthalmids were
previously thought to be closely related to and derived from the
Citharidae (Hubbs, 1945). This hypothesis was based on certain
symplesiomorphies, e.g., the low degree of fusion of the gill
membranes and the presence of vomerine teeth. The Scoph-
thalmidae show hypural pattern 6 and are thus members of the
bothoid group.
Paralichthyidae.— 'Norman (1934) basically defined the
Paralichthyinae (=Paralichthyidae with modifications) on ex-
ternal pelvic-fin morphology and vertebral structure (absence
of transverse apophyses). The group was supposed to have the
ventral fins nearly symmetrical in position and base lengths, or
the ocular fin on the midventral line and its base slightly ex-
tended anteriorly. Symmetries in ventral-fin position and base
lengths are plesiomorphic for the order and bothoids. Norman's
paralichthyid genera with an ocular ventral fin on the midventral
line and its base extended anteriorly are bothids (i.e., Trichop-
setta, Engyophrys, Taeniopsetta, Monolene. Perissias).
Amaoka (1969) presented a more thorough, detailed defini-
tion of the family. However, many or most of the character
states he used appear to be plesiomorphic for bothoid fishes
(i.e., those defined by hypural pattern 6). A second limitation
of Amaoka's work on this group is that it was limited to three
genera {Paralichthys, Pseudorhombus. Tarphops). An important
change in Norman's classification was made by Amaoka when
he removed Taeniopsetta from the Paralichthyidae and placed
it in the Bothidae. Hensley (1977) and Futch (1977) did the
same for Monolene, Engyophrys, and Trichopsetta.
We have now examined some characters in the remaining
Paralichthyidae, and additional changes are required in the com-
position of this group. In a survey of caudal-fin structure, it was
found that Thysanopsetta and Tephrinectes show the most prim-
itive type of hypural pattern (1; Fig. 363 upper). These two
genera are much more primitive than expressed in the current
classification and definitely do not belong to the bothoid group.
Within the remaining Paralichthyidae another group is dis-
cemable. This is composed of Cyclopsetta, Syacium, Citharich-
thys, and Etropus, i.e., the Cyclopsetta group. States for two
complexes of characters, ventral-fin morphology and urinary-
papilla position, are unique to this group and interpreted as
synapomorphic. Arrangement of caudal-fin rays in the Cyclop-
setta group is also unique and probably apomorphic (Fig. 364
middle). All species have 1 7 caudal rays, none of which are
supported by preural, neural or haemal spines. It should also
be noted that the fifth hypural has fused with an epural. This
fusion has been observed in larval development (Tucker. 1982;
Ahlstrom, pers. observ.). However, fusion of the fifth hypural
and one or more epurals has apparently occurred several times
in pleuronectiforms, possibly including the bothoids (e.g., see
Fig. 360). A detailed analysis of relationships between the Cy-
clopsetta group and other bothoids is not possible here. How-
ever, some character states may indicate a close relationship
with bothids (absence of first neural spine, presence of vertebral
transverse apophyses).
Amaoka (1 969) and one of us (Ahlstrom) recognized another
group within the Paralichthyidae composed of Pseudorhombus,
Tarphops. and Cephalopsetla. i.e.. the Pseudorhombus group.
We interpret these genera as more specialized in certain char-
acters than most other members of the family. Species of this
group usually have a total caudal ray count of 17, the epural is
fused to the fifth hypural, and they lack a splinter ray on the
ventralmost caudal-fin ray. With the exceptions noted above
(the primitive non-bothoid genera Tephrinectes and Thysan-
opsetta and the Cyclopsetta group), the remaining paralichthyids
of the Regan-Norman classification (what we are calling the
Paralichthys group) have the apparently plesiomorphic states
of 1 8 caudal rays, at least one free epural (except in one species
of Hippoglossina (Sumida et al., 1979)), and a splinter ray on
the ventralmost caudal-fin ray (Fig. 364 lower). The splinter ray
is probably a remnant of a ray lost through fusion with an
adjacent ray (Okiyama, 1974). The Pseudorhombus group may
be definable by synapomorphies but a detailed analysis has not
been done.
After removal of the bothids (Trichopsetta, Engyophrys, Tae-
niopsetta, Monolene, Perissias) and the primitive non-bothoid
genera (Tephrinectes, Thysanopsetta), recognition of the Cy-
clopsetta group as monophyletic. and recognition of the Pseu-
dorhombus group as possibly monophyletic, few of the original
paralichthyid genera remain. We have been referring to these
as the Paralichthys group (Ancylopsetta, Gastropsetta, Hippo-
glossina, Lioglossina, Paralichthys, Verecundum, and Xystreu-
rys). At least most of the character states known for these re-
maining genera are plesiomorphic for the order (e.g., symmetrical
ventral-fin states) or for bothoids (e.g.. usual presence of at least
one free epural). The Paralichthys group is probably not mono-
phyletic.
Bothidae.— Norman (1934) defined the Bothinae (=Bothidae
with modifications) on the basis of a high degree of ventral-fin
asymmetry and the presence of vertebral transverse apophyses.
The ocular ventral fin was said to be on the midventral line with
its base extending anteriorly to the urohyal. Norman excluded
Taeniopsetta, Engyophrys, Trichopsetta, Monolene, and Peris-
sias from this group because the base of the ocular ventral fin,
although on the midventral line and somewhat longer than that
of the blind side, does not extend to the urohyal.
Amaoka (1969) examined many bothid genera and redefined
the family using more characters. Most of the characters stressed
by Amaoka have now been examined in other bothoids. These
are discussed below:
Ventral-fin asymmetry.— In bothids the ocular fin base is on the
midventral line, elongated, and has its origin anteriorly placed
relative to the base of the blind fin. Within the bothoids this
combination of states appears to be derived and unique.
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
683
Preorbital on blind side.— This bone is absent in the Bothidae.
It appears to be present in all other bothoids (Pleuronectinae
not examined for this character). Based on this comparison, we
interpret the loss of this bone a derived state within the bothoids
defining the family Bothidae.
Infraorbital bones of the ocular side. — All bothids have an ocu-
lar preorbital bone but lack the remainder of the series. The
presence or absence of the ocular preorbital has not been sur-
veyed in most bothoid groups. However, an ocular infraorbital
lateral line is present in most bothoids. In addition to the Both-
idae, it is missing in Brachypleura and the Cyclopsetta group.
Intermuscular bones.— We interpret the presence of at least two
of the series of these bones (myorhabdoi) as a derived state
unique to and defining the Bothidae.
First neural spine.— Although the first neural arch is present,
the neural spine is missing in the Bothidae. It is present in all
other bothoids except the Cyclopsetta group.
Vertebral transverse apophyses.— All bothids have two pairs of
transverse apophyses on most vertebrae. As previously dis-
cussed, how to interpret these on the pieuronectiform level and
within the bothoid group is questionable. Within the bothoids
well-developed and very similar structures occur only in the
Bothidae and Scophthalmidae. Very small transverse apophyses
also occur in the Cyclopsetta group.
Based on these characters, the Bothidae appear to be mono-
phyletic and definable by synapomorphies in at least three char-
acters or complexes: (1) loss of the preorbital on the blind side;
(2) presence of myorhabdoi; and (3) asymmetrical states of ven-
tral-fin morphology.
Since Amaoka's ( 1 969) work, we have examined the remain-
ing genera not examined by him that have been considered
bothids (i.e., Graminatobothus. Lophonectes, Pelecanichlhys,
Mancopsetta). All of these except Mancopsetta are bothids.
Mancopsetta exhibits the following character states: ( 1 ) hypural
pattern 1, i.e., the most primitive type; (2) presence of pleural
and epipleural ribs, but no myorhabdoi or other intermuscular
bones in the caudal region; (3) at least one free epural (none in
adult bothids); (4) anus on midventral line (clearly on blind side
in bothids); (5) no vertebral transverse apophyses; and (6) seven
rays in the ocular ventral fin, 5-7 in that of the blind side (six
in both fins in bothids). These are all characters in which Man-
copsetta differs from the Bothidae. Due to the primitive hypural
pattern, it is not a bothoid (see Rhombosoleinae).
Amaoka (1969) analyzed intergeneric relationships of Jap-
anese bothids. However, his analysis was eclectic and did not
include all genera (i.e., Engyophrys. Trichopsetta, Monolene,
Perissias. Graminatobothus. Lophonectes. and Pelecanichlhys
were not examined). He recognized two subfamilies, the Tae-
niopsettinae and Bothinae. He erected the first subfamily for
Taeniopsetta. Hensley (1977), Futch (1977), Evseenko (1977,
1981), and Amaoka (1979) implied that Engyophrys and Tri-
chopsetta should be included in the Taeniopsettinae. This was
done on the basis of larval characters and ventral-fin morphol-
ogy. Most of the slates used to define the Taeniopsettinae were
considered by Amaoka (1969) to be plesiomorphic at the family
level. Three characters were emphasized: (1) degree of anterior
extension of the base of the ocular ventral fin; (2) shape of the
ventral (sciatic) area of the urohyal; and (3) number of suborbital
bones on the blind side. In the taeniopsettines, the origin of the
blind ventral fin is at the same transverse level as the second
ray of the ocular ventral fin, i.e., the base of the ocular fin is
only slightly elongated. In the Bothinae, extension of the base
of the ocular fin is greater and the origin of the blind fin is on
the same transverse level as the third or fourth ray of the ocular
fin. Obviously, the taeniopsettine state here is the more plesio-
morphic. Engyophrys, Trichopsetta, Monolene, and Perissias
show this state. Taeniopsetta has a broad, truncate margin on
the sciatic part of the urohyal. In bothines, this area of the
urohyal is pointed. Amaoka (1969) clearly showed that the ple-
siomorphic state for bothoids is closer to the condition shown
in taeniopsettines. Engyophrys, Trichopsetta, and Perissias show
the taeniopsettine condition, Monolene the bothine state.
Amaoka (1969) noted an apparent trend among bothoids in
reduction of the number of suborbital bones of the blind side.
This reduction may have occurred in several bothoid groups
and interpretation of this character is not clear. Thus, infraor-
bital counts for bothoids are as follows (preorbital -t- suborbit-
als): Scophthalmidae 1+5; Brachypleura 1 + 0; Paralichthys
group 1 + 4-5; Pseudorhombus group 1 + 5-7; Cyclopsetta
group 1 + 5-6; and Bothidae 0 + 3-5. Pleuronectines were not
examined for this character. The most common count in both-
oids other than bothids is 1 + 5-7. Thus, there is some evidence
that the basal or plesiomorphic count for bothids may be five
suborbitals on the blind side. Among bothids this count appar-
ently occurs only in Taeniopsetta and Pelecanichthys. Engyo-
phrys, Trichopsetta, Perissias, and Monolene have three sub-
orbitals on the blind side. In summary, there is good evidence,
at least for the first two characters discussed above, that the
Taeniopsettinae show states that are plesiomorphic for the fam-
ily and may not be monophyletic.
Pleuronectidae.— Norman (1934) considered this family to be
one of the "higher" flatfish groups, i.e., those with a mono-
morphic optic chiasma. Hubbs (1945) basically followed this
interpretation, but showed that two of Norman's pleuronectid
genera, Brachypleura and Lepidoblepharon. possessed some
primitive states not shown in other pleuronectids. These two
genera were removed by Hubbs and placed in his family Cith-
aridae.
Norman (1934) defined the Pleuronectidae as being dextral
and having eggs without oil globules. Basic to his concept of
this family were the assumptions that all members were mono-
morphic in regard to the optic chiasma and that nearly all species
were discriminately dextral. He divided the family into five
subfamilies. All members of the Poecilopsettinae, Paralichthod-
inae, Samarinae. and Rhombosoleinae, as presently interpreted,
are discriminately dextral, i.e., sinistral individuals occur so
rarely in any one species that they can be considered anomalies.
Most species of Pleuronectinae are also discriminately dextral.
The few exceptions have probably returned to indiscriminate
ocular asymmetry secondarily (Hubbs and Hubbs, 1945). We
have no reason to doubt Norman's or Hubbs' assumption that
the Pleuronectinae have a monomorphic optic chiasma. How-
ever, as previously discussed, there are no data showing this for
the other pleuronectid subfamilies. Uniting these groups in the
family Pleuronectidae appears to have rested only on ocular
asymmetry. We have surveyed these subfamilies for various
684
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
characters and are confident that the Pleuronectidae as currently
defined are not monophyletic. In fact, four of the pleuronectid
subfamilies are not bothoids as we define the group. However,
what the true relationships of these groups are is unknown. We
discuss these subfamilies individually:
Poecilopsettinae.— We have examined radiographs of speci-
mens ofPoecilopsetta and Nematops. These genera have hypural
pattern 1, at least one free epural, 20 caudal rays, and what
appears to be a haemal-arch remnant on the parhypural. The
caudal structure here is primitive compared to the bothoids and
these fishes do not belong to that group. Poecilopsettines are
poorly known and character states defining the group or relating
it to others have not been investigated.
?ar?i\ic)M\iod.mdie:. —Paralichthodes algoensis has hypural pat-
tern 1 (Ahlstrom, pers. observ.) and does not belong to the
bothoid group. Its relationships to other groups are unknown.
Samarinae.— Since Hubbs' (1945) removal oi Brachypleura and
Lepidoblepharon from this group, it has been composed of Sa-
maris and Samanscus. We have not done a detailed study of
these genera, but some characters we have examined are worthy
of note: ( 1 ) These genera show a unique hypural pattern (5; Fig.
364 upper). We interpret this pattern as derived relative to
pattern 1 and as indicative that the group is monophyletic. Using
this pattern to relate the group is more difficult: however, one
of us (Ahlstrom) noted that in late-stage larvae of Samanscus.
hypural pattern 1 is present, and fusions resulting in pattern 5
must occur very late in development. This is evidence that
pattern 5 may have evolved directly from pattern 1 and does
not represent a modification of the bothoid pattern 6. (2) Sa-
marines are the only pleuronectiforms known other than the
Bothidae to have intermuscular bones, although they do not
have the two series of myorhabdoi as found in bothids. We have
not done a detailed study of these bones in samarines, but they
appear very similar to the epimerals, epicentrals. and hypom-
erals of bothids. (3) Samarines, cynoglossids, and soleines have
an anterior pair of well-developed transverse apophyses on many
vertebrae. Two pairs of these structures are found in the Both-
idae and Scophthalmidae. (4) The Samannae, Soleoidei, and
Mancopselta lack postcleithra, at least in adults. How to inter-
pret these last three character states is open to question. Are
three of the series of intermuscular bones homologous in sa-
marines and bothids? Are the anterior vertebral transverse
apophyses homologous between all of the groups? Do some of
these character states indicate a close relationship between sa-
marines and some soleoids (i.e., cynoglossids and soleines)? Our
tentative hypothesis is that the samarines are a line that is at
least independent from the bothoids. Here we are obviously
stressing caudal characters. The corollary of this is that we are
interpreting similarities between samarines and bothoids in in-
termuscular bones and vertebral transverse apophyses as hom-
oplasies.
Rhombosoleinae. — The main character states used by Norman
(1926, 1934) to define this subfamily were the high degree of
asymmetry in the ventral fins and the absence of pectoral radials.
The ocular ventral fin is on the midventral line and its base is
considerably extended. The blind ventral fin is short based or
missing. Another interesting characteristic of this group is that
several genera show high numbers of fin rays in the ocular ven-
tral fin. There is a great deal of morphological diversity in rhom-
bosoleines. Some genera appear fairly generalized in many char-
acters (Oncopterus. Psammodiscus. Rhombosolea. Azygopus. and
Pelotretis); others are more specialized (Colistiuin. Peltorham-
phits, and Ammotretis). Many of the specializations in the latter
genera are similar to those in some soleoids. This has been
interpreted as parallel evolution (Norman, 1934; Hubbs, 1945).
Norman apparently had some doubts about aligning this group
with the Pleuronectinae. He realized that Parker's (1903) ex-
amination of one specimen of Oncopterus darwinii in his survey
of optic chiasmata did not prove the group to be monomorphic
in this character. This group has still not been studied in detail.
It may be monophyletic, but its relationship to other flatfishes
is unknown.
We have examined the caudal skeleton of all rhombosoleine
genera except Psammodiscus. They show hypural patterns 1 and
4 (Fig. 363 upper and lower). Assuming the group is monophy-
letic, there are two implications here: (1) The primitive pleu-
ronectiform hypural pattern 1 is also plesiomorphic for the
Rhombosoleinae, and the derived pattern 4 arose within the
group independently from the same pattern in the Soleinae,
Cynoglossidae, and Eucttharus. (2) The Rhombosoleinae are
not bothoids and should not be aligned with the Pleuronectinae.
The possibility has recently become apparent that Mancop-
setla may be most closely related to the Rhombosoleinae. All
known specimens of Mancopselta are sinistral and it has been
considered a bothid. However, it shares certain character states
with at least some rhombosoleines. This genus has ventral-fin
ray counts of 7 on the ocular side and 5-7 on the blind side.
Although not strictly limited to the rhombosoleines, these high
counts, at least in the fin of the ocular side, are characteristic of
at least four rhombosoleine genera. The eyes are densely scaled
in Mancopsetta and in Azygopus and Pelotretis. However, scaled
eyes are found in some genera of other groups also (e.g., some
pleuronectines). Andnashev (1960) and Penrith (1965) have
both remarked on a fleshy lip-like structure which overhangs
the anterior end of the upper jaw in .Mancopsetta. One of the
soleoid-type characteristics exhibited by the more specialized
rhombosoleines is the dorsal fin originating in a rostral hook
that overhangs the mouth. In the more generalized genera, there
is no rostral hook and the dorsal fin originates at some posterior
position. In at least one of these generalized genera (Azygopus.
the only one examined for this character) there is a fleshy struc-
ture (possibly a precursor to the rostral hook?) overhanging the
anterior end of the upper jaw which is very similar to that in
Mancopsetta. Obviously more comparative work needs to be
done here. However, it is possible that Mancopsetta and the
Rhombosoleinae may form a monophyletic group with an in-
discriminately dextral or sinistral common ancestor.
Pleuronectinae. — Norman (1934) stressed two character states
in defining this subfamily: ( 1 ) lateral line well developed on both
sides of the body; and (2) olfactory laminae parallel (except in
Atheresthes), without rachis. A well-developed lateral line on
both sides of the body is plesiomorphic for the order and both-
oids. We have not examined olfactory laminae or attempted to
analyze distributions of states for the character.
We have shown that the Pleuronectidae is probably not mono-
phyletic, due to the inclusion of the four non-bothoid subfam-
ilies. The subfamily Pleuronectinae is the only bothoid group
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
685
in Norman's Pleuronectidae. Members of this subfamily are
dextral or apparently secondarily indiscriminate (Hubbs and
Hubbs, 1945). They apparently have a monomorphic optic
chiasma. However, most character states which species of this
subfamily share appear to be plesiomorphic for the order or
bothoids. e.g.. symmetrical or nearly symmetrical ventral-fin
placement and fin-base lengths, anus on or close to the mid-
ventral line. We have examined the caudal osteology of about
half of the pleuronectine genera. All have the bothoid hypural
pattern (6) and one or possibly two free epurals. We have found
no synapomorphies in the caudal fin for this group.
Larval characters
In the previous discussion, many doubts were raised con-
cerning pleuronectiform interrelationships as expressed in the
Regan-Norman model. Unfortunately, larvae for many of these
groups are unknown. A second problem is that surveys for many
characters where larvae are known have been incomplete and
inconsistent. Most descriptive larval research has dealt with
characters useful for identification and has not involved com-
parative work of sufficient detail to determine homologous states.
Such work is sorely needed before distnbutions of homologous
states can be determined for many characters.
Below is a list and discussion of certain characters and com-
plexes. Selection of these was based mainly on the amount of
available information.
Preopercidar spines. — The presence of preopercular spines ap-
pears to be plesiomorphic for the order and some pleuronecti-
form groups. This is based on the observation that the slate is
widespread among flatfish and percomorph larvae.
Neurocranial spines. —Spines occur in some regions of the neu-
rocranium in some pleuronectiform larvae. Most of these are
said to occur in the otic or frontal regions. However, determining
homologies here is difficult due to a general lack of detailed
osteological study of the bones carrying these spines. Spines in
the otic and frontal regions appear to be of two types. One of
these is where spines are associated with neurocranial ridge
systems. These are known for larvae of achirines (Houde et al.,
1970; Futch et al., 1972), some scophthalmids (Jones, 1972),
and some pleuronectines (Pertseva-Ostroumova, 1961). In the
second type, spines occur singly or in small groups but are not
part of a pronounced ridge. These have been said to occur on
various bones of the otic region (epiotics, autosphenotics, au-
topterotics) or on the frontals. Tucker (1982) was not able to
determine the origin of such spines in the larvae of Citharichthys
and Etropus and referred to them as frontal-sphenotic spines.
Although thorough studies are needed before neurocranial spines
can be used to infer or test pleuronectiform interrelationships,
certain patterns are noteworthy: (1) Spines that are not part of
some pronounced ridge system appear to be limited to some
bothoids (some species of the Paralichthys group, Cyclopsetta
group, Pseudorhombus group. Scophthalmidae, Pleuronectinae,
and Bothidae). (2) Within the Bothidae, only the larvae of En-
gyophrys. Taeniopsetla, and Trichopselta (Taeniopsettinae; lar-
vae of Perissias are unknown) are known to have otic spines
(Amaoka. 1979). In these genera, the spines are on the same
bones (epiotics and autosphenotics) and are probably homol-
ogous. (3) Within the Cyclopsetta group, a relatively well-de-
veloped otic or frontal spine occurs in Cyclopsetta and Syacium
(Aboussouan, 1968b; Gutherz, 1970; Ahlstrom, 1971; Futch
and HoflT, 1971; Evseenko. 1979), while series of small spines
occur in Citharichlfiys and Etropus (Tucker. 1982).
Urohyal, basipterygial. and cleithral spines. Spines on these
bones are limited to certain genera of the Bothidae. Thus, they
are considered apomorphic at the pleuronectiform and bothoid
levels of universality.
Early-forming elongated dorsal-fin rays. — The presence of elon-
gated dorsal-fin rays in pleuronectiform larvae has been exten-
sively and justifiably used for identification purposes. However,
use of these structures for phylogenetic interpretations is pres-
ently difficult and generally premature. There are several reasons
for this. Surveys for these characters are inadequate, since larvae
for many groups are unknown. Characters and character states
have never been adequately defined to allow proper compari-
sons to be made. The only pattern here that is clear and phy-
logenetically interpretable is the state in bothids. All species of
this family for which larvae are known show elongation of only
the second dorsal-fin ray. This state is known only in this family
and thus appears to be apomorphic within the order and both-
oids.
Early-forming elongated ventral- fin rari.— Ocular ventral-fin
rays which are elongated relative to those of the blind side are
limited to certain species of the Cyclopsetta group. Due to the
restricted occurrence of these, they are probably apomorphic
for the order and bothoids. However, within the Cyclopsetta
group, the distribution of elongated ocular ventral-fin rays does
not conform to generic groups based on adult morphology. At
least one species of cynoglossid is known to have elongated rays
in the ventral fin of the blind side (Kyle, 1913; Padoa, 1956k).
Size at metamorphosis. — MosX flatfishes metamorphose in the
size range of ca. 10-25 mm. When size at metamorphosis has
been discussed in regard to evolution in pleuronectiforms, the
usual hypothesis has been that certain species and groups have
evolved mechanisms for prolonging larval life for greater dis-
persal, and others have actually shortened larval life for re-
cruitment to limited habitats (Amaoka, 1979; Moser, 1981).
There are several implications in this hypothesis that are rele-
vant here; (1) There is some size range for transformation that
is plesiomorphic for the order. This is usually implied to be ca.
10-25 mm because most pleuronectiforms metamorphose in
this range. (2) Metamorphosis at markedly smaller (e.g., Achir-
inae) or larger (e.g., Bothidae, some pleuronectines) sizes are
derived states. (3) According to the Regan-Norman model, pro-
longed larval development must have developed independently
in several lines. Although metamorphosis at large sizes is most
common in bothids, it is also known for some Pleuronectinae,
the Poecilopsettinae, some species of the Cyclopsetta group, and
some cynoglossids.
Size at metamorphosis is an important character for larval
identification, but its use for inferring phylogenetic relationships
in most instances is premature. Exceptions may exist in the
Bothidae, where the extremely long premetamorphic lengths
exhibited by some genera are probably apomorphic within the
family and can be used for phylogenetic information.
686
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Relative time of caudal-fin formation.— In most known larvae
of flatfishes and other teleosts, formation of the caudal fin pre-
cedes or occurs with that of the dorsal and anal fins. The only
exceptions known in pleuronectiforms are the cynoglossids. In
this family, the caudal fin does not develop until the dorsal and
anal fins are nearly completely developed. This pattern of de-
velopment is considered apomorphic in pleuronectiforms.
Eye migration and dorsal-fin position at tnetamorphosis. — Eye
migration has been observed in some flatfish groups. In the
Psettodidae, Pleuronectinae, Paralichthyidae (excluding the Cy-
clopsetta group), Scophthalmidae, and apparently some Sole-
idae, the first ray of the dorsal fin is above or posterior to the
eyes. At metamorphosis, the migrating eye crosses anterior to
the dorsal-fin origin. These types of eye migration and dorsal-
fin position appear to be plesiomorphic for the order. Several
derived states for these characters occur. In at least one species
of cynoglossid, a fleshy rostral beak is formed anterior to the
dorsal-fin origin. Eye migration takes place between the rostral
beak and the interorbital region. In some soleids, the dorsal-fin
origin projects above the snout and the eye migrates between
this projection and the neurocranium. In the Bothidae, the dor-
sal fin is anterior to the eye and attached to the ethmoid region.
During migration, the eye goes between the base of the dorsal
fin and the ethmoid region. A path for the migrating eye is
created by detachment of the anterior section of the dorsal fin
from the ethmoid region so that a narrow slit is formed, or some
tissue in the path of the migrating eye is absorbed. A very similar
type of eye migration occurs in some species of the Cyclopsetta
group. However, in other members of this group, the eye mi-
grates around the dorsal-fin origin (Gutherz, 1970; Tucker, 1982).
Phylogenetic information provided by
larval characters
Although larvae of some critical groups are unknown or poor-
ly known, some comments about phylogenetic relationships can
be made in regard to groups where our knowledge is on a higher
level.
Bothoids. —Spines in the otic or frontal regions of the neuro-
cranium which are isolated or in small clusters appear to be
limited to various groups of bothoids. If these spines prove to
be homologous between these groups, they may be apomorphic
within the order. In this interpretation, they would be primitive
for bothoids and lost in various lines.
Paralichthyidae.— As discussed in the section on adult charac-
ters, this family as currently interpreted is polyphyletic due to
the inclusion of Tephrinectes and Thysanopsetta. We do not
consider these genera bothoids as defined by the caudal-fin com-
plex. Their larvae are unknown.
We have interpreted the Cyclopsetta group as monophyletic
based on some adult character states which are probably apo-
morphic. Although larvae of this group show certain states which
appear to be apomorphic within bothoids (e.g., elongated left
ventral-fin rays), not all species in this group show these.
The Pseudorhombus group is possibly definable by adult syn-
apomorphies. In larvae of this group, we see no character states
that are presently interpretable with certainty as synapomor-
phies.
In examining adult characters of the Paralichthys group, it
appeared likely that this group had no synapomorphies. Larvae
tend to support this. They show the following character states
which appear to be plesiomorphic for the order: (1) presence of
preopercular spines; (2) origin of the dorsal fin behind the eyes;
(3) metamorphosis in a size range of 7.5-14.2 mm; and (4) eye
migration anterior to the dorsal fin. In addition, at least some
species show the following states which may prove to be ple-
siomorphic at least within the bothoids: (1) four or five elon-
gated, early-forming dorsal-fin rays; and (2) presence of otic
spines.
Bothidae. — 'With the exclusion of Mancopsetta and inclusion of
Perissias, this family is definable by adult synapomorphies. Lar-
vae of the Bothidae are probably better known than for any
other family of flatfishes. However, larvae of many genera are
still unknown (i.e., Parabothus, Asterorhombus, Tosarhombus.
Neolaeops. Japonolaeops, and Perissias). Amaoka (1979) re-
viewed larval characters of most genera for which larvae are
known. Known bothid larvae show the following character states
which are interpreted as synapomorphies: (1) metamorphosis
at a relatively large size (ca. 15-120 mm); (2) eye migration
below the dorsal fin; (3) dorsal-fin origin anterior to eyes just
prior to metamorphosis; (4) elongated, early-forming second
dorsal-fin ray; and (5) lack of preopercular spines.
Larvae of some bothid genera have various combinations of
otic-region, urohyal, cleithral, and basipterygial spines. It is
tempting to use the presence of these spines to define bothid
groups, and therefore, assume that they are apomorphic within
the family. Spines in the otic region within the Bothidae are
limited to the Taeniopsettinae as presently defined. However,
spines in this region occur in other bothoid groups. Although
sufficient comparative osteological work has not been done to
show that these spines are homologous between taeniopsettines
and other bothoids, use of these spines to infer close relation-
ships between Engyophrys. Taeniopsetta. and Trichopsetta is
questionable. Urohyal, cleithral, and basipterygial spines are
known only from larvae of nine bothid genera. They occur in
various combinations inter- and intragenerically. Amaoka (1969)
presented a model of intergeneric relationships for Japanese
bothids based on adult characters. Occurrence of these larval
spines is scattered among the bothid lines hypothesized by
Amaoka. This could indicate two possibilities: ( 1 ) the spines are
apomorphic within the family, and Amaoka's model is incor-
rect; or (2) Amaoka's model is correct and the spines are ple-
siomorphic within the family and have been lost in several lines.
Two major problems exist with Amaoka's phytogeny based on
adult characters; it was constructed using eclectic methods and
it did not include all genera. Interpretation of urohyal, basip-
terygial, and cleithral spines should await a cladistic analysis of
bothid interrelationships based on adult characters.
Pleuronectidae. — Based on adult characters, we interpret this
family as polyphyletic. Larvae of the four non-bothoid subfam-
ilies are poorly known, and hence, of little aid in determining
relationships of these groups. However, there are certain simi-
larities in general body morphology between the few known
samarine and poecilopsettine larvae. In regard to the Pleuro-
nectinae, many adult states that are shared are plesiomorphic
for pleuronectiforms or bothoids. This also appears to be true
for most larval characters. The position of the dorsal-fin origin
(posterior to the eyes) and the type of eye migration (anterior
to the dorsal-fin origin) are plesiomorphic for the order. Some
pleuronectine larvae have preopercular spines, which again, are
HENSLEY AND AHLSTROM: PLEURONECTIFORMES
687
probably plesiomorphic for flatfishes. Some genera show spines
in the otic region of the neurocranium; these are possibly ple-
siomorphic for bothoids. All known pleuronectine larvae lack
elongated dorsal-fin rays. However, this state is not limited to
this group and a phylogenetic interpretation of it would be pre-
mature. In short, at present, we know of no character states that
are unique to pleuronectine larvae or that can confidently be
interpreted as apomorphic.
Egg characters
Except in certain groups, eggs of flatfishes are still too poorly
known to be of much value in phylogenetic studies. One char-
acter of pleuronectiform eggs was used by Regan (1910) and
Norman (1934) to interpret phylogeny, the presence of one oil
globule in bothid eggs to separate them from those of pleuro-
nectids which lack oil globules. We now have more information
about the occurrence of oil globules in flatfish eggs, and the
distribution of these character states is not exactly that predicted
by the Regan-Norman model (preceding article, this volume).
The obvious pattern here is that bothoids have 0-1 and soleoids,
rhombosoleines, and Mancopsetta multiple oil globules. There
are published exceptions to this. Watson and Leis (1974) iden-
tified three types of eggs with multiple oil globules as those of
bothids. However, these authors expressed some doubt about
the identifications of at least two of these egg types. These eggs
are probably some other group (poecilopsettines or samarines?).
Brownell (1979) identified some eggs which lacked oil globules
as the soleid Heteromycteris capensis. This is the only soleid we
are aware of that lacks multiple oil globules.
It is probably premature to use the oil-globule character for
phylogenetic information until eggs from other groups are known.
However, it is interesting and possibly significant that the so-
leoids, rhombosoleines, and Mancopsetta are so sharply sepa-
rable from the bothoids in this character. One oil globule appears
to be the most common state in the eggs of percomorph fishes
(based on accounts in Watson and Leis, 1974; Russell, 1976;
Fritzsche, 1978; Hardy, 1978b; Johnson, 1978; and Brownell,
1979). This may indicate that this state is plesiomorphic for
pleuronectiforms. Corollaries of this would be that oil globules
were lost in most pleuronectines, and multiple oil globules de-
veloped in a line leading to the soleoids, rhombosoleines, and
Mancopsetta.
(D.A.H.) Department of Marine Sciences, University of
Puerto Rico, Mayaguez, Puerto Rico 00708.
LITERATURE CITED
Abe, T. 1954. New, rare or uncommon fishes from Japanese waters.
IV. Records of rare fishes of Ihe families Lophotidae, Nomeidae,
Icosteidae. Jpn. J. Ichthyol. 3:90-95.
. 1957. Notes on fishes from the stomachs of whales taken in
the Antarctic. Sci. Rep. Whales Res. Inst. (Tokyo) No. 12:225-233.
, andS. Kaji. 1972. A. record of Oreoso?na atlanlicuni(OTeoso-
matidae, Zeiformes, Teleostei) from Tasman Sea. U.O. No. 8:5-7.
Able, K. W. 1984. Variation m spawning site preference of the mum-
michog, Fundulus heleroclilus. Copeia 1984:522-525.
. MS. Comments on the distribution and reproductive season-
ality of snailfishes and lumpfishes in the estuary and Gulf of St.
Lawrence.
, J. D. Felley and D. Powers. MS. Taxonomy of Bermuda
Fundulus (Pisces: Fundulidae).
, D. F. Markle AND M. P. Fahay. MS. Development of larval
liparidine snailfishes from the western North Atlantic.
, and D. E. McAllister. 1980. Revision of the snailfish genus
Liparis from Arctic Canada. Can. Bull. Fish. Aquat. Sci. No. 208.
, and J. A. Mlisick. 1976. Life history, ecology and behavior
oi Lipans inquilmus (Pisces: Cycloptendae) associated with the sea
scallop, Placopecten magetlanicus. Fish. Bull. U.S. 74:409-421.
Aboussouan, A. 1964. Contribution a I'etude des oeufs et larves pe-
lagiques des poissons teleosteens dans la Golfe de Marseille. Rec.
Trav. Stn. Mar. Endoume Fac. Sci. Mars. 32(48):87-l 73.
. 1965. Oeufs et larves de teleosteens de I'ouest africain. 1.
Acanthurus monroviae Sleind. Bull. Inst. Fondam. Afr. Noire, Ser.
A Sci. Nat. 27:1183-1187.
. 1966a. Sur quelques larves pelagiques de teleosteens recoltees
a Tulear. Reel. Trav. Stn. Mar. Endoume-Mars. Suppl. 5:149-164.
. 1966b. Sur quelques larves pelagiques de teleosteens recoltees
a Tulear. Cent. Oceanogr. Fac. Sci. Mars. 1966:147-159.
. 1966c. Oeufs et larves de teleosteens de I'ouest africain. III.
Larves de Monacanthus hispidus (L.) et de Balistes forcipalus Gm.
Bull. Inst. Fondam. Afr. Noire, Ser. A 28:276-282.
. 1966d. Oeufs et larves de teleosteens de fouest africain. IV.
Galeoides polydactylus (Vahl) [Polynemidae]. Bull. Inst. Fondam.
Afr. Noire, Ser. A Sci. Nat. 28:1037-1040.
. 1967. Oeufs et larves de teleosteens de I'ouest africain. V.
Caranx rhonchus GeofTroy Saint-Hilaire (Carangidae). Affinites avec
Trachurus trecae Cadenat. Bull. Inst. Fondam. Afr. Noire, Ser. A
Sci. Nat. 29:1039-1050.
. 1968a. Oeufs et larves de teleosteens de I'ouest africain. VI.
Larves de Chloroscombrus chrysurus (Linne) et de Blepharius crin-
itis (Mitchill) (Carangidae). Bull. Inst. Fondam. Afr. Noire, Ser. A
Sci. Nat. 30:226-237.
. 1968b. Oeufs et larves de teleosteens de I'ouest africain VII.
Larves de Syacium giiineensis (Blkr.) [Bothidae]. Bull. Inst. Fon-
dam. Afr. Noire, Ser. A Sci. Nat. 30:1 188-1 197.
— . 1968c. Oeufs et larves de teleosteens de I'ouest africain. VIll.
Larves de Bregmaceros macclellandi Thompson (Bregmaceroti-
dae). Bull. Inst. Fondam. Afr. Noire, Ser. A Sci. Nat. 30:1590-
1602.
— . 1 969. Sur une petite collection de larves de teleosteens recoltee
au large du Bresil (Campagne "Calypso" 1962). Vie Milieu, Ser. A
Biol. Mar. 20:595-610.
— . 1972a. Oeufs et larves de teleosteens de I'ouest africain. X.
Larves d'Ophidioidei (Oligopus. Ophidian et Carapus) et de Per-
coidei (Pseudupeneus). Bull. Inst. Fondam. Afr. Noire, Ser. A Sci.
Nat. 34:169-178.
— . 1972b. Oeufs et larves de teleosteens de I'ouest africain. XI.
Larves serraniformes. Bull. Inst. Fondam. Afr. Noire, Ser. A Sci.
Nat. 34:485-502.
— . 1972c. Oeufs et larves de teleosteens de I'ouest africain. Xll.
Les larves d'Heterosomata recoltees aux environs de I'ile de Goree
(Senegal). Bull. Inst. Fondam. Afr. Noire, Ser. A Sci. Nat. 34:974-
1003.
. 1975. Oeufs et larves de teleosteens de I'ouest africain. XIII.
Contribution a I'identification des larves de Carangidae. Bull. Inst.
Fondam. Afr. Noire, Ser. A Sci. Nat. 37:899-938.
. 1980. Description d"unelarvegeante"rubaniforme,"attribuee
au genre Brotulataenia (Pisces, Gadiformes, Ophidioidei, Ophi-
diidae) et recoltee a Test des nouvelles Hebrides. Cybium, 3rd Ser.,
(IO):51-64.
Adams, J. A. 1960. A contribution to the biology and postlarval de-
velopment of the sargassum fish, Histno hisino (Linnaeus), with a
discussion of the Sargassum complex. Bull. Mar. Sci. Gulf Caribb.
10:55-82.
Agassiz, A. 1882. On the young stages of some osseous fishes. Proc.
Am. Acad. Arts Sci. 16(Pt. 3):27 1-303.
, AND C. O. Whitman. 1885. The development of osseous
fishes. I. The pelagic stages of young fishes. Mem. Mus. Comp.
Zool. Harv. 14(1): 1-56.
Ahlstrom, E. H. 1943. Studies on the Pacific pilchard or sardine
(Sardinops caerulea). 4. Influence of temperature on the rate of
development of pilchard eggs in nature. U.S. Fish Wildl. Serv. SSRF
23.
. 1956. Eggs and larvae of anchovy, jack mackerel, and Pacific
mackerel, p. 33-42. In: Progress Rep., Calif Coop. Oceanic Fish.
Invest., 1 April 1955-30 June 1956, Sacramento.
. 1965. Kinds and abundance of fishes in the California Current
region based on egg and larval surveys. Calif Coop. Oceanic Fish.
Invest. Rep. 10:31-52.
. 1968. Review of "Development of fishes of the Chesapeake
Bay region, an atlas of egg, larval and juvenile stages. Part 1".
Copeia 1968:648-651.
. 1969. Remarkable movements of oil globules in eggs of bathy-
lagid smelts during embryonic development. J. Mar. Biol. Assoc.
India 11:206-217.
. 1971. Kinds and abundance of fish larvae in the eastern trop-
ical Pacific, based on collections made on EASTROPAC 1. Fish.
Bull. U.S. 69:3-77.
. 1972a. Kinds and abundance of fish larvae in the eastern
tropical Pacific on the second multivessel EASTROPAC survey,
and observations on the annual cycle of larval abundance. Fish.
Bull. U.S. 70:1153-1242.
. 1972b. Distributional atlas of fish larvae in the California
Current region: six common mesopelagic fishes. Calif Coop. Oceanic
Fish. Invest. Atlas 17.
— . 1974. The diverse patterns of metamorphosis in gonostomatid
fishes - an aid to classification, p. 659-674. In: The early life history
offish. J. H. S. Blaxter (ed.). Springer-Verlag, Berlin.
— . 1 976. Maintenance of quality in fish eggs and larvae collected
during plankton hauls, p. 313-318. //;. Zooplankton fixation and
preservation. H. F. Steedman (ed). LInesco Press, Paris.
— , and O. P. Ball. 1954. Description of eggs and larvae of jack
mackerel ( Trachurus symmelricus) and distribution and abundance
of larvae in 1950 and 1951. Fish. Bull. U.S. 56:209-245.
— , J. L. Butler and B. Y. Sumida. 1976. Pelagic stromateoid
fishes (Pisces, Perciformes) of the eastern Pacific: kinds, distribu-
tions and early life histories and observations on five of these from
the northwest Atlantic. Bull. Mar. Sci. 26:285-402.
— , AND R. C. Counts. 1955. Eggs and larvae of the Pacific hake
Merlucaus produclus. Fish. Bull. U.S. 56:295-329.
— , and . 1958. Development and distribution of iinci-
guerria lucetia and related species in the eastern Pacific. U.S. Fish
Wildl. Serv. Fish. Bull. 58:363-416.
— , AND H. G. Moser. 1969. A new gonostomatid fish from the
tropical eastern Pacific. Copeia 1969:493-500.
— , AND . 1975. Distributional atlas offish larvae in the
California Current region: Flatfishes, 1955 through 1960. Calif
Coop. Oceanic Fish. Invest. Atlas 23.
— , AND . 1976. Eggs and larvae of fishes and their role in
systematic investigations and in fisheries. Rev. Trav. Inst. Peches
Mant. 40:379-398.
688
LITERATURE CITED
689
.AND . 1980. Characters useful in identification of pelagic
manne fish eggs. Calif. Coop. Oceanic Fish. Invest. Rep. 21:121-
131.
, AND . 1981. Systematics and development of early life
history stages of marine fishes: achievements during the past cen-
tury, present status and suggestions for the future. Rapp. P. V. Reun.
Cons. Int. Explor. Mer 178:541-546.
— , AND M. J. O'TooLE. 1976. Development and distri-
bution of larvae and early juveniles of the commercial lantemfish,
Lampanyctodes hecloris (Gijnther). off the west coast of southern
Africa with a discussion of phylogenetic relationships of the genus.
Bull. South. Calif Acad. Sci. 75:138-152.
— , AND E. G. Stevens. 1977. Report of neuslon (surface) collec-
tions made on an extended CALCOFI cruise dunng May 1972.
Calif Coop. Oceanic Fish. Invest. Rep. 18:167-180.
-, AND B. Y. Slimida. MS. Early life history of eastern Pacific
carangids. NMFS, U Jolla, CA 92038.
Akatsu. S., Y. OciASAWARA andF. Yasuda. 1977. Spawning behavior
and development of eggs and larvae of the striped fingerfish, Mono-
daaylus sebac. Jpn. J. Ichthyol. 23:208-214.
Akazaki. M.. H. Nakajima, H. Kawahara and S. Takamatsii. 1 976.
Embryonic development and metamorphosis after hatching in the
sharksucker. Echeneis naucrates. Jpn. J. Ichthyol. 23:153-159. (In
Japanese, English abstract.).
Akihito, p. 1963. On the scapula of gobiid fishes. Jpn. J. Ichthyol.
11:1-26.
. 1967. Additional research on the scapula of gobiid fishes. Jpn.
J. Ichthyol. 14:167-182.
. 1969. A systematic examination of the gobiid fishes based on
the mesoptcrygoid, postcleithra, branchiostegals. pelvic fins, scap-
ula, and suborbital. Jpn. J. Ichthyol. 16:93-1 14.
, ANI5 K. Mecjuro. 1981. A gobiid fish belonging to the genus
Hetereteolris coWecleA in Japan. Jpn. J. Ichthyol. 28:329-339.
Alcock, a. 1889. Natural history notes from H. M. Indian Manne
Survey Steamer 'Investigator,' Commander Alfred Carpenter. R.N.,
D.S.O., Commanding. — No. 13. On the bathybial fishes of the Bay
of Bengal and neighbouring waters, obtained dunng the seasons
1885-1889. Ann. Mag. Nat. Hist., Ser. 6. 4:450-461.
Alderdice, D. F.. AND C. R. Forrester. 1971. Effects of salinity and
temperature on embryonic development of petrale sole (Eopsetia
jordani). J. Fish. Res. Board Can. 28:727-744.
Alexander, E. C 1961. A contribution to the life history, biology
and geographical distribution of the bonefish, Alhtilii vulpes (Lin-
naeus). Dana-Rep. Carlsberg Found. 53.
Allen, G. H. 1968. Fecundity of the brown ragfish, Icosleus aenig-
maticus Lockington, from northern California. Calif Fish Game
54:207-214.
Allen, G. R. 1972. The anemonefishes. Their classification and bi-
ology. T.F.H. Publications, Inc.. Neptune City, NJ.
. 1975a. Damselfishes of the south seas. T.F.H. Publ.. Inc.,
Neptune City, NJ.
. 1975b. The biology and taxonomy of the cardinalfish 5p/iae-
ramia orbicularis (Piices: Apogomdae). J. R. Soc. West. Aust. 58(3):
86-92.
. 1980. A genenc classification of the rainbowfishes (family
Melanotaeniidae). Rec. West. Aust. Mus. 8:449-490.
. 1981. Polynemidae. In: FAO species identification sheets.
Eastern Central .Atlantic, Fishing Areas 34, 37. W. Fischer and W.
B. Scott (eds.). FAO, Rome.
, and H. J. Cross. 1980. Descriptions of five new rainbowfishes
(Melanotaeniidae) from New Guinea. Rec. West. Aust. Mus. 8:
377-396.
, and R. H. KtUTFR. 1976. A review of the plesiopid fish genus
Assesor. with descriptions of two new species. Rec. West. Aust.
Mus. 4(3):20l-215.
Allen.J. M., J. H.S. Blaxter andE. J. Denton. 1976. The functional
anatomy and development of the swim bladder— inner ear— lateral
line system in herring and sprat. J. Mar. Biol. Assoc. U.K. 56:471-
486.
Allen, L. G. 1979. Lar\a\ deve\opmeM of Gohiesox rhessodon (Go-
biesocidae) with notes on the larva of Rimicola muscarum. Fish.
Bull. U.S. 77:300-304.
. and J. M. Ilg. 1983. Larval development of the northern
clingflsh, Gobiesox maeandricus. Copeia 1983:551-554.
Altukhov, K, A. 1978. On the reproduction and abundance of the
lesser sand lance, Ammodyles marinus. in the White Sea. J. Ich-
thyol. (Engl. Transl. Vopr. Ikhtiol.) 18:560-567.
. 1979. A contribution to the biology of non-commercial and
rare fish species from the White Sea in the early stages of devel-
opment. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 19(6):105-1 15.
Alvarez-Lajonchere, L. L. 1975. Estudio sistematico de Afugil hra-
siliensis. Mugil gaimardianus, y Mugil curema. Cienc. Ser. 8 Invest.
Mar. (Havana) No. 14:1-18.
Amaoka, K. 1964. Development and growth of the sinistral flounder,
Bothus myriaster (TemmmcV. and Schlegel) found in the Indian and
Pacific Oceans. Bull. Misaki Mar. Biol. Inst. Kyoto LIniv. 5:1 1-29.
. 1969. Studies on the sinistral flounders found in the waters
around Japan. Taxonomy, anatomy, and phylogeny. J. Shimono-
seki Univ. Fish. 18:65-340.
. 1970. Studies on the larvae and juveniles of the sinistral floun-
ders. I. Taeniopseua ocellata (Gunther). Jpn. J. Ichthyol. 17:95-
104.
. 1971. Studies on the larvae and juveniles of the sinistral floun-
ders. II. Chascanopsella lugiibris. Jpn. J. Ichthyol. 18:25-32.
. 1972a. Studies on the larvae and juveniles of the sinistral
flounders. III. Laeops kitaharae. Jpn. J. Ichthyol. 19:154-165.
. 1972b. Osteology and relationships of the citharid flatfish
Brachypleura novaezeetandiae. Jpn. J. Ichthyol. 19:263-273.
. 1973. Studies on the larvae and juveniles of the sinistral floun-
ders. IV. Arnoglossus japonicus. Jpn. J. Ichthyol. 20:145-156.
. 1974. Studies on the larvae and juveniles of the sinistral floun-
ders. V. Arnoglossus tenuis. Jpn. J. Ichthyol. 21:153-157.
. 1976. Studies on the larvae and juveniles of the sinistral floun-
ders. VI. Psetlma iijimae. P tosana. and P gigantea. Jpn. J. Ich-
thyol. 22:201-206.
. 1979. Phylogeny and larval morphology of pleuronectiform
fishes. (Psettodidae. Citharidae, Paralichthyidae and Bothidae).
Kaiyo Kagaku 1 1(2):I00-I 10. (English transl., 1981. Edited by J.
R. Dunn, Translated by A. Shiga. Natl. Mar. Fish. Serv., Northwest
and Alaska Fish. Center, Seattle).
, M. ToKOSHiMA and T. Inada. 1977. New records of the
stichaeid fish Ascoldia variegata knipowitschi and the zoarcid fish
Puzanovia rubra from Japan. Jpn. J. Ichthyol. 24:91-97.
Ambrose. D. A., J. L. Butler, H. G. Moser, B. Y. Sumida, E. M.
Sandknop and E. G. Stevens. 1983. Description of the larvae
of the cusk-eels Ophidian scrippsae and Chilara taylori (Ophidi-
idae). Calif Coop. Oceanic Fish. Invest. Rep. 24:226-234.
Andersen, N. C 1984. Genera and subfamilies of the family No-
totheniidae (Pisces, Perciformes) from the antarctic and subant-
arctic. Steenstrupia 10:1-34.
, AND J. C HiREAU. 1979. Proposition pour une nouvelle clas-
sification des Nototheniinae (Pisces, Perciformes, Nototheniidae).
Cybium, 3rd Ser., 3(6):47-53.
Anderson, M. E. 1981. Aspects of the natural history of the midwater
fish Lycodapus mandibulans (Zoarcidae) in Monterey Ba\. Cali-
fornia. Pac. Sci. 34:181-194.
. 1982. Revision of the fish genera Gymnetus Reinhardt and
Gymnetopsis Soldatov (Zoarcidae), with two new species and com-
parative osteology of Gymnclus vindis. Natl. Mus. Nat. Sci. (Ot-
tawa) Publ. Zool. 17:1-76.
. 1984. On the anatomy and phylogeny of Zoarcidae (Teleostei:
Perciformes). LInpubl. PhD dissertation. Coll. William and Mary,
Williamsburg, VA.
, andG. M. Cauliet. 1974. Crab and snailfish commensalism
in Monterey Bay. Underwater Nat. 8(3):29-31.
, AND C L. Hi BBS. 1981. Redescription and osteology of the
northeastern Pacific fish Derepodichlhys alepidotus (Zoarcidae). Co-
peia 1981:341-352.
690
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Anderson, W. D., Jr.. and P. C. Heemstra. 1980. Two new species
of western Atlantic Anlhias (Pisces: Serranidae), redescription of
A. asperilinguis and review of Holanthias martinicensis. Copeia
1980:72-87.
Anderson, W. W. 1957. Early development, spawning, growth, and
occurrence of the silver mullet (Mugil curema) along the south
Atlantic coast of the United States. Fish. Bull. U.S. 57:397-414.
, J. W. Gehringer and F. H. Berry. 1966. Family Synodon-
tidae. p. 30-102. In: Fishes of the western North Atlantic. G. W.
Mead (ed.). Mem. Sears Found. Mar. Res. No. 1 (Pt. 5).
Anderton, T. 1907. Observations on New Zealand fishes, etc., made
at the Portobello marine fish-hatchery. Trans. NZ Inst. 39:477-
496.
Andrew, W. 1959. Textbook of comparative histology. Oxford Univ.
Press, New York.
Andriashev, a. p. 1938. Review of the genus Krusensterniella Schmidt
(Pisces, Zoarcidae), with description of a new species from the Japan
Sea. Bull. Far East. Br. Acad. Sci. USSR 32:1 17-121.
. 1952. A new fish of the eelpout family (Pisces. Zoarcidae)
from the Benng Sea. Zool. Inst. Acad. Nauk SSSR 12:415-417.
. 1954. Fishes of the northern seas of the USSR. Tr. Zool. Inst.
Akad. Nauk SSSR, No. 53. (In Russian, translated by Isr. Program
for Sci. Transl., Jerusalem, 1 964. Available from NTIS, Springfield.
VA; OTS 63-11 160.
. 1955a. A fish newto thefaunaoftheUSSR— fn/ep/iliin/epM
zonifer (LocV.) Pisces, Anoplopomidae] from the Kamchatkan waters
of the Pacific Ocean. Vopr. Ikhtiol. 4:3-9.
. 1955b. Observations on the eel-like likods (Lr(v«i:-/)Wv5 Gill
(Pisces, Zoarcidae) and related forms) in the seas of the USSR and
adjacent waters. Tr. Zool. Inst. Acad. Nauk SSSR 18:349-384.
. 1959. On the number of vertebrae and certain other osteo-
logical characters of Antarctic fishes. Vopr. Ikhtiol. 12:3-7. (In
Russian.)
. 1960. Families of fishes new to the Antarctic. Communique
3. Pelagic young of flatfish (Pisces, Bothidae) off the Antarctic coast.
Zool. Zh. 39:1056-1061.
. 1962. Bathypelagic fishes of the Antarctic. 1. Family Myc-
tophidae. Akad. Nauk SSSR Issled. Fauny Morei l(9):2l6-294. (In
Russian; Bur. Comm. Fish. Ichthyol. Lab. Transl. 29.)
. 1965. A general review of the Antarctic fish fauna. Monogr.
Biol. 15:491-550.
— . 1975. A new ultra-abyssal fish, Notolipans kurchatovi gen. et
sp. n. (Liparidae) from the South Orkney Trench (.Antarctica). Tr.
Inst. Okeanol. Akad. Nauk SSSR 103:313-319.
— , and a. V. Neelov. 1976. Geniolipans lindhergi, gen. et sp.
n.— a new fish of the family of sea snails (Liparidae) from the bathyal
depths of Antarctica, p. 68-77. In: Zoogeography and systematics
of fishes. V. M. Korovina (ed). Zool. Inst. Acad. Sci. USSR, Len-
ingrad.
— , , and V. P. Prirodina. 1977. On methods of study of
the morphology and systematics of the fish family of sea snails
(Liparidae). Zool. Zh. 50:141-147.
AoYAMA, T. 1959. On the egg and larval stages of Liparis tanakae
(Gilbert and Burke). Bull. Seikai Reg. Fish. Res. Lab. 18:69-73.
, and N. Sotogak.1. 1955. On the development of the egg of
Nemtplerus virgalus (Houttuyn). Jpn. J. Ichthyol. 6:130-132.
Aprieto, V. L. 1974. Early development of five carangid fishes of the
Gulf of Mexico and the south Atlantic coast of the United States.
Fish. Bull. U.S. 72:415-443.
Apsangikar, D. K. 1953. The systematic position of 5/TO»Mre«i «/-
ger. J. Univ. Bombay, N.S. 21B(5):41-50.
Apstein, C. 1909. Die Bestimmungdes Alters pelagisch lebenderSee-
fischeier. Mitt. Dtsch. Seefischereiver 25:364-373.
Arata, G. F.. Jr. 1954. A contribution to the life history of the
swordfish. Xiphias gladius Linnaeus, from the South Atlantic coast
of the United States and the Gulf of Mexico. Bull. Mar. Sci. Gulf
Caribb. 4:183-243.
Arbault, S., and N. Boutin. 1968a. Ichthyoplancton.Oeufset larves
de poissons teleosteens dans le golfe de Gascogne en 1964. Rev.
Trav. Inst. Peches Marit. 32:413-476.
, and . 1968b. Ichthyoplancton. Oeufs et larves de pois-
sons teleosteens dans le Golfe Gascogne en 1964. Rev. Trav. Inst.
Peches Marit. 32:477-506.
, AND . 1968c. Ichthyoplancton. Oeufs et larves de pois-
sons teleosteens dans le Golfe de Gascogne en 1965-1966 (Oeufs
et larves). Rev. Trav. Inst. Peches Marit. 33:181-202.
Armstrong, P. B. 1962. Stages in the development of /c7a/Hn« /ic/)-
ulosus. Syracuse University Press, Syracuse, New York.
, AND J. S. Child. 1965. Stages in the normal development of
Fundutus heteroditus. Biol. Bull. (Woods Hole) 128(2):143-168.
Arnold, D.C. 1956. A systematic revision of the teleost family Carap-
idae (Percomorphi. Blennioidea), with descriptions of two new
species. Bull. Bnt. Mus. (Nat. Hist.) Zool. 4:245-307.
Aronovich, T. M., S. I. DoROSHEv, L. V. Spectorova and V. M.
Makhotin. 1975. Egg incubation and larval rearing of navaga
(Eleginus navaga Pall) polar cod (Boreogadus saida Lepechin) and
arctic flounder (Liopsetta glacialis Poll.) in the laboratory. Aqua-
culture 6:233-242.
AsANo, H. 1962. Studies on the congrid eels of Japan. Bull. Misaki
Mar. Biol. Inst. Kyoto Univ. 1:1-143.
Astakhov, D. A. 1980. Cranial lateral-line canals of the beloniform
fishes (Beloniformes, Osteichthyes). Tr. Inst. Okeanol. Akad. Nauk
SSR 97:174-21 1. (In Russian, English abstract.)
Auer, N. a. (ed.). 1982. Identification of larval fishes of the Great
Lake Basin with emphasis on the Lake Michigan Drainage. Great
Lakes Fish. Comm. Spec. Publ. 82-3.
Auten, M., AND F. Bonhomme. 1980. Elements de systematique bio-
chimique chez les Mugilides de Mediterranee. Biochim. Syst. Ecol.
8(3):305-308.
AziiMA, H. 1975. Spawning the red-bellied piranha, Serrasabnus nat-
tereri. Trop. Fish Hobby. 23(ll):4-6. 8-12, 14-15, 96-97.
. 1982. Spawning silver. Trop. Fish Hobby. 30(7):70-73.
Backhouse, G. N., and R. W. V.anner. 1978. Observations on the
biology of the dwarf galaxiid Galaxiella pusilla (Mack) (Pisces:
Galaxiidae). Victorian Nat. 95:128-132.
Backl'S, R. H.. J. E. Craddock. R. L. Haedrich and B. H. Robison.
1977. Atlantic mesopelagic zoogeography, p. 266-287. In: Fishes
of the western North Atlantic. R. H. Gibbs, Jr. (ed.). Mem. Sears
Found. Mar. Res. No. 1 (Pt. 7).
, G. W. Mead, R. L. Haedrich AND A. W. Ebeling. 1965. The
mesopelagic fishes collected during Cruise 1 7 of the R/V Chain,
with a method for analyzing faunal transects. Bull. Mus. Comp.
Zool. 134:139-157.
Badcock, J. R. 1982. A new species of the deep-sea fish genus Cy-
clothone Goode and Bean (Stomiatoidei, Gonostomatidae) from
the tropical Atlantic. J. Fish Biol. 20:197-21 1.
, .\ND N. R. Merrett. 1972. On Argynpnus allanlicus. Maul
1952 (Pisces, Stomiatoidei), with a description of post-larval forms.
J. Fish Biol. 4:277-287.
, AND . 1976. Midwater fishes in the eastern North At-
lantic. I. Vertical distribution and associated biology in 30°N, 23°W,
with developmental notes on certain myctophids. Prog. Oceanogr.
7:3-58.
Bagarinao.T., AND J. R. Hunter. 1983. The visual feeding threshold
and action spectrum of northern anchovy (Engraulis mordax G.)
larvae. Calif Coop. Oceanic Fish. Invest. Rep. 24:245-254.
Bailey. C. F. 1973. Lipids of the fertilized egg and adult brain of
Fundulus heteroditus. J. Exp. Zool. 185:265-276.
Bailey, K. M. 1982. The early life history of the Pacific hake Mer-
lucaus productus. Fish. Bull. U.S. 80:589-598.
, AND R. S. Batty. 1983. A laboratory study of predation by
Aurelia aurita on larval herring (Clupea harengits). Experimental
observations compared with model predictions. Mar. Biol. 72:295-
301.
Bailey, R. G. 1972. Observations on the biology oi Nolhobranchius
guentheri (PfefCer) (Cyprinodontidae), an annual fish from the coast-
al region of East Afnca. Afr. J. Trop. Hydrobiol. Fish. 2:33-47.
Bainbridoe, V. 1961. The early life history of the bonga Ethmalosa
dorsalis (Cuvier and Valenciennes). J. Cons. Int. Explor. Mer 26:
347-353.
LITERATURE CITED
691
. 1962. The larvae of fe/ZoMi^/a vora.v Gunther (Clupeidae) in
Sierra Leone coastal waters. Bull. Inst. Fr. Afr. Noire, Ser. A 24:
262-269.
Baird, R. C. 1971. The systematics. distribution, and zoogeography
of the marine hatchetfishes (Family Stemoptychidae). Bull. Mus.
Comp. Zool. 142:1-128.
, D. F. Wilson, R. C. Beckett .-".nd T. L. Hopkins. 1974. Dia-
phus taamngi Norman, the principal component of a shallow sound-
scattering layer in the Cariaco Trench, Venezuela. J. Mar. Res. 32:
301-312.
, AND D. M. MiLLiKEN. 1973. Observations on Breg-
maceros nectabanus Whitley in the anoxic, sulfurous waters of the
Cariaco Trench. Deep-Sea Res. 20:503-504.
Baker, A. N. 1973. Spawning and development of the New Zealand
sprat Spratlus antipodum. Zool. Publ. Victoria Univ. Wellington
62:1-12.
Balakrishnan, K. p. 1961. Fish eggs and larvae collected by the
research vessel "Conch" during 1957-61. 1. Larvae of C'v«og/oMi«
semifasciatus Day. Bull. Cent. Res. Inst. Univ. Kerala India Ser. C
Nat. Sci. 8:131-139.
. 1963. Fish eggs and larvae collected by the research vessel
"Conch": I. Larvae oi Arnoglossus tapaemosoma Blkr., Bolhus
ocellatus Agassiz, Laeops giienlhen Ale, Solea ovala Rich., and
Cynoglossus monopus Blkr. Bull. Dep. Mar. Biol. Oceanogr. Univ.
Kerala 1:81-96.
, AND C. B. L. Devi. 1974. Larvae of some Hal fishes from a
tropical estuary, p. 677-684. In: The early life history offish. J. H.
S. Blaxter (ed.). Springer- Verlag, New York.
Balbontin, P., I. Campodonico AND L. Guzman. 1979. Descripcion
de huevos y larvas de especies de Careproctus comensales de Par-
alomis granutosay Lithodes anlarctica. An. Inst. Patagonia 10:235-
243.
, AND R. Perez. 1980. Descripcion de los estados larvales de
Normanichthys crocken Clark (Perciformes: Normanichthyidae) del
area de Valparaiso. Rev. Biol. Mar. 17(l):81-95.
Ballard, J. 1970. Aquanum briefs. Bull. S. Afr. Assoc. Mar. Biol.
Res. 8:31-34.
Balon, E. K. 1956. Development of the huchen [Hiicho huclio (L.)]
during the period of exogenous feeding after hatching.
Pornohospodarstvo 3:433-455. (In Slovak.)
. 1959. Die Entwicklung der Texas— Cichlide (//cnc/»/n'i cvfl-
nogultalus Baird et Girard) nach dem Schliipfen. Zool. Anz. 162
(ll-l2):339-355.
. 1975a. Reproductive guilds of fishes: a proposal and defini-
tion. J. Fish. Res. Board Can. 32:821-864.
. 1975b. Terminology of intervals in fish development. J. Fish.
Res. Board Can. 32:1663-1670.
. 1977. Early onxogeny of Labeotropheus Ahl. 1927 (Mbuna,
Cichlidae, Lake Malawi), with a discussion on advanced protective
styles in fish reproduction and development. Environ. Biol. Fishes
2:147-176.
(ed.). 1980. Charrs: salmonid fishes of the genus Salvelinus.
Perspectives in vertebrate science I. Dr. W, Junk, The Hague.
. 1981a. Additions and amendments to the classification of
reproductive styles in fishes. Environ. Biol. Fishes 6:377-389.
1981b. Saltatory processes and altncial to precocial forms in
the ontogeny of fishes. Am. Zool. 21:573-596.
BANARESfU, P. 1964. Pisces. Osteichthyes. Fauna Repub. Pop. Rom.
(Bucuresti) 13.
Banister, K. E. 1967. The anatomy and classification of the order
Gasterosteiformes (Pisces). Unpubl. PhD dissertation. Univ. New-
castle upon Tyne.
Bapat, S. V. 1955. A preliminary study of the pelagic fish eggs and
larvae of the Gulf of Mannar and the Palk Bay. Indian J. Fish.
2:231-255.
, and R. R. Prasad. 1952. On some developmental stages of
Caranx kalla Cuv. & Val. J. Bombay Nat. Hist. Soc. 51:111-115.
Baracloi'gh, W. E. 1964. Contribution of the marine life history of
Thateichlhys paaficus. J. Fish. Res. Board Can. 21:1333-1337.
Barans. C. A., AND A. C. Barans. 1972. Eggs and early larval stages
of the spotted hake, Urophycis regius. Copeia 1972:188-190.
Barham, E. G. 1966. An unusual pelagic flatfish observed and pho-
tographed from a diving saucer. Copeia 1966:865-867.
Barkman, R. C. 1978. The use of otolith growth rings to age young
Atlantic silversides, Menidia menidia. Trans. .Am. Fish. Soc. 107:
790-792.
Barnard, K, H. 1926. The eggs of the pilot-fish (A'aiicTa/M c/wcvor).
Nature (Lond.) 118:228.
. 1927. A monograph of the marine fishes of South Africa. Pari
II. Ann. S. Afr. Mus. 21:419-1065.
Barnhart, P. S. 1927. Pelagic fish eggs of La Jolla, California. Bull.
Scripps Inst. Oceanogr. Univ. Calif. 1:91-92.
. 1932. Notes on the habits, eggs, and young of some fishes of
southern California. Bull. Scripps Inst. Oceanogr. Univ. Calif. 3(4):
87-99.
Barrington, E. J. W. 1937. The structure and development of the
tail in the plaice (Pleiironecles platessa) and the cod (Gadus mor-
rhua). Q. J. Microsc. Sci. 79:447-469.
Barsukov, V. V. 1959. The wolffish (Anarhichadidae). Fauna SSSR
Moscow N.S. 73.(Transl. by Smithson. Inst., 1972. Available from:
NTIS, Spnngfield, VA.: TT 67-59074.)
. 1973. The species composition of the genus //e&o/e^wi (Se-
bastinae. Scorpaenidae: Pisces) and a description of a new species.
J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 13:161-167.
. 1981. A brief review of the subfamily Sebastinae. J. Ichthyol.
(Engl. Transl. Vopr. Ikhtiol.) 21(l):l-26.
Bartlett, M. R., and R. L. Haedrich. 1966. Techniques in the
radiography of fishes. Trans. Am. Fish. Soc. 95:99-101.
Bassot, J.-M. 1966. On the comparative morphology of some lu-
minous organs, p. 557-610. In: Bioluminescence in Progress. F.
Johnson and Y. Haneda (eds.). Princeton Univ. Press. Princeton,
N.J.
. 1971. Structure and evolution of the light organs of the sto-
miatoidei (Abstract), p. 253-254. In: The ocean world. M. Uda
(ed.). Proc. Joint Oceanogr. Assembly.
Bath, H. 1973. Wiederbeschreibung und neuer Nachweis von Tnpre-
rygion melanurus (Guichenot) (Pisces: Blennioidea: Tripterygi-
idae). Senckenb. Biol. 54:47-56.
. 1976. Revision der Blenniini (Pisces: Blenniidae). Senckenb.
Biol. 57:167-234.
. 1978. Erganzungen zur Revision der Blenniini mit zwei neun
Gattungen (Pisces: Blenniidae). Senckenb. Biol. 59:183-190.
Badchot, M. L. 1959. Etude des larves leptocephales du group Lep-
lochephalus lanceolalus Stromman et identification a la famille des
Serrivomeridae. Dana-Rep. Carlsberg Found. 48:1-148.
Baugh, T. M. 1980. Spawning and hatching the Olympic mudminnow
{NoYumhra hubbsi) in aquaria. J. Aquaric. l(3):63-64.
Baxter, J. L. 1 960. A study of the yellowtail Seriola dorsalis (Gill).
Calif. Dep. Fish Game Fish Bull. 1 10.
Beck, A. D., D. A. Benotson AND W. H. Howell. 1980. International
study on Anemia. V. Nutritional value of five geographical strains
of Anemia: effects on survival and growth of larval Atlantic sil-
verside, Menidia menidia. p. 249-259. In: The brine shrimp .Ir-
lemia. Vol. 3. Ecology, cultunng, use in aquaculture. G. Persoone,
P. Sorgeloos, O. Roels and E. Jaspers (eds.). Universa Press. Wet-
teren, Belgium.
Beebe, W. 1932. Nineteen new species and four post-larval deepsea
fish. Zoologica (N.Y.) 13:47-107.
. 1933a. Deep-sea fishes of the Bermuda Oceanographic Ex-
peditions. No. 2. Family Alepocephalidae. Zoologica (N.Y.) 1 6(2):
15-93.
. 1933b. Deep-sea fishes of the Bermuda Oceanographic Ex-
peditions. No. 3. Family Argentinidae. Zoologica (N.Y.) I6(3):97-
147.
. 1934. Deep-sea fishes of the Bermuda Oceanographic Expe-
ditions. Family Idiacanthidae. Zoologica (N.Y.) 16(4): 149-241.
, AND J. Crane. 1939. Deep-sea fishes of the Bermuda Ocean-
ographic Expeditions. Family Melanostomiatidae. Zoologica (N.Y.)
24(2):65-238.
692
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Behnke, R. J. 1968. A new subgenus and species of trout. Salmo
(Platysalmo) platycephalus. from southcentral Turkey, with com-
ments on the classification of the subfamily salmoninae. Mitt. Hamb.
Zool. Mus. Inst. 66.
Bekker, V. E. 1963. Taxonomy and distribution of Tarletonheania
(Myctophidae, Pisces). Tr. Inst. Okeanol. Akad Nauk SSSR 62:
145-163. (In Russian.)
. 1964. Slendertailed myctophids (genera Loweina. Tarleton-
heania. Gonichthys and Centrobranchus) of the Pacific and Indian
Oceans, systematics and distribution. Tr. Inst. Okeanol. Akad. Nauk
SSSR 73:11-75 (In Russian.)
. 1965. Lantemfishes of the genus A/i',?cip/!i/»; (Myctophidae,
Pisces), systematics and distribution. Tr. Inst. Okeanol. Akad. Nauk
SSSR 80:62-103. (In Russian; Bur. Comm. Fish. Ichthyol. Lab.
Transl. 45.)
. 1967. The lantemfishes (Myctophidae) from the "Peter Le-
bedev" Atlantic Expeditions 1961-1964. Tr. Inst. Okeanol. Akad.
Nauk SSSR 84:84-124. (In Russian.)
, AND O. D. BoRODLiLiNA. 1968. Lantcmfishes of the genus
Ceratoscopelus Gunther, systematics and distribution. Vopr. Ikh-
tiol. 8(5):779-798. (In Russian.)
, AND . 1978. "Myclophum asperum" species— group
with description of a new species, and Myclophum selenops TSning
(Myctophidae, Osteichthys), taxonomy and distribution. Tr. Inst.
Okeanol. Akad. Nauk SSSR 111:108-128. (In Russian.)
Belanger, L. F. 1973. The skeletal tissues, p. 175-215. /«. Histology.
R. O. Creep and L. Weiss (eds.). McGraw Hill, New York.
Bell, M. A. 1976. Evolution of phenotypic diversity in Gasterosleus
aculeatus superspecies on the Pacific Coast of North America. Syst.
Zool. 25:211-227.
Belman, B. W., AND M. E. Anderson. 1979. Aquanum observations
on feeding by Melanostigma pammelas (Pisces. Zoarcidae). Copeia
1979:365-369.
Belyanina, T. N. 1974. Materials on the development, systematics
and distribution of fish of the family Bregmacerotidae. Tr. Inst.
Okeanol. Akad. Nauk. SSSR 96:143-188.
. 1977. Materials on development of C/ia«/(0(^Mi fishes (Chau-
liodontidae, Pisces). Tr. Inst. Okeanol. Akad. Nauk. SSSR 109:
113-132.
. 1 980. Codlets (Bregmacerotidae, Osteichthyes) of the Carib-
bean Sea and the Gulf of Mexico. J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 20(0:138-141.
. 1981. The larvae of some rare mesopelagic fishes from the
Caribbean and the Gulf of Mexico. J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 21(l):82-95.
. 1982a. Larvaeof the family Scopelarchidae (collections of the
Institute of Oceanology Academy of Sciences of USSR). Tr. Inst.
Okeanol. Akad. Nauk SSSR. 1 18:133-150. (In Russian.)
. 1982b. Larvae of the midwater fishes in the western tropical
Pacific Ocean and the seas of the Indo-Australian Archipelago. Tr.
Inst. Okeanol. Akad. Nauk SSSR 1 18:5-42.
, AND N. V. KovALEvsicAVA. 1979. Materials of development
and distribution of myctophid larvae (Family Myctophidae) of the
sub-antarctic waters of the Australian-New Zealand waters. Tr. Inst.
Okeanol. Akad. Nauk SSSR 106:69-96.
, AND G. R. Lopes. 1974. The occurrence oi Bregmaceros nec-
tabanus ( Pisces, Bregmacerotidae) in the Caribbean Sea. J. Ichthyol.
(Engl. Transl. Vopr. Ikhtiol.) 14:447-449.
Bennett, H. S., A. D. Wyrick, S. W. Lee and J. H. McNeil. 1976.
Science and art in preparing tissue embedded in plastic for light
microscopy, with special reference to glycol methacrylate, glass
knives and simple stains. Stain Technol. 51:71-97.
Bensam, p. 1 967. The embryonic and early larval development of the
long-finned herring, Opisthoplenis lardoore (Cuvier). J. Mar. Biol.
Assoc. India 9:76-83.
. 1970. Notes on the eggs, larvae and juveniles of the Indian
Sprat, Sardinella jussieui (Lacepede). Indian J. Fish. 13:219-231.
Ben-tlivia, A. 1975. Mugilid fishes of the Red Sea with a key to the
Mediterranean and Red Sea species. Bamidgeh 27(1): 14-20.
Benzie, V. 1968a. Stages in the normal development of Galaxias
maculatus allenualus Uenyn^). N.Z. J. Mar. Freshwater Res. 2:606-
627.
. 1968b. The life history of Galaxias vulgaris Stokell, with a
comparison with G. maculalus allenualus. N.Z. J. Mar. Freshwater
Res. 2:628-653.
Berg, L. S. 1940. Classification of fishes, both recent and fossil. Tr.
Zool. Inst. Akad. Nauk SSSR 5(2). (In Russian.)
, 1947. Classification of fishes both recent and fossil. J. W.
Edwards, Ann Arbor, Michigan.
. 1948. Freshwater fishes of the U.S.S.R. and adjacent countries.
Vol. I. Acad. Sci. U.S.S.R. Zool. Inst.. Keys to the fauna of the
U.S.S.R., No. 27. (Translated from the Russian, 1962, Israel Pro-
gram for Scientific Translations.)
Berkeley, S. A., and E. D. HouDE. 1978. Biology of two exploited
species of halfbeaks, Hemiramphus brasiliensis and H. balao from
southeast Flonda. Bull. Mar. Sci. 28:624-644.
Berra, T. M. 1982. Life history of the Australian grayling Pro/o/rortM
maraena (Salmoniformes: Prototroctidae) in the Tambo River,
Victoria. Copeia 1982:795-805.
Berrien, P. L., N. A. Naplin and M. R. Pennington. 1981. Atlantic
mackerel Scomber scombrus. egg production and spawning popu-
lation estimates for 1977 in the Gulf of Maine, Georges Bank and
Middle Atlantic Bight. Rapp. P. V. Reun. Cons. Int. Explor. Mer
178:279-288.
Berr'i, F. H. 1958. A new species of fish from the western North
Atlantic, Dikellorhynchus tropidnlepis. and relationships of the gen-
era Dikellorhynchus and Malacanthus. Copeia 1958:1 16-125.
. 1959a. Boarfishes of the genus .-)«r;,?o^i/a of the western At-
lantic. Bull. Ha. State Mus. Biol. Sci. 4(7):205-250.
. 1959b. Young jack crevalles (Caraw.v species) off the south-
eastern Atlantic coast of the United States. U.S. Fish Wildl. .Serv.
Fish. Bull. 59:417-535.
. 1960. Scaleandscutedevelopment of the carangid fish Cara^i.v
crysos (Mitchell). Q. J. Ha. Acad. Sci. 23:59-66.
. 1 964a. Aspects of the development of the upper jaw bones in
teleosts. Copeia 1964:375-384.
. 1964b. Review and emendation of Family Clupeidae. Copeia
1964:720-730.
. 1969. f/agato Wp(/j/7Mtoa (Pisces: Carangidae): morphology
of the fins and other characters. Copeia 1969:454-463.
— , and W. W. Anderson. 1961. Stargazer fishes from the western
North Atlantic (family Uranoscopidae). Proc. U.S. Natl. Mus. 1 12
(3448):563-586.
— , and W. J. Baldwin. 1966. Triggerfish (Balistidae) of the east-
em Pacific. Proc. Calif Acad. Sci. 34:429-474.
— , and I. Barrett. 1963. Gillraker analysis and speciation in
the thread herring genus Opisthonema. Inter-Am. Trop. Tuna
Comm. Bull. 7:112-190.
— , AND H. C. Perkins. 1966. Survey of pelagic fishes of the
California Current area. U.S. Fish Wildl. Serv. Fish. Bull. 65:625-
682.
— , AND W. J. Richards. 1973. Characters useful to the study of
larval fishes. Mid. Atl. Coast. Fish. Cent. Tech. Pap. 1:48-65.
— , and C. R. Robins. 1967. Macnstiella perlucens. a new clu-
peiform fish from the Gulf of Mexico. Copeia 1967:46-50.
— , AND W. F. Smith-Vaniz. 1978. Carangidae. //!. FAO species
identification sheets for fisheries purposes. Western Central Atlan-
tic, Fishing Area 31. Vol. I. W. Fischer (ed.). FAO, Rome.
— , AND L. E. VoGELE. 1961. Filefishes (Monacanthidae) of the
western north Atlantic. U.S. Fish Wildl. Serv. Fish. Bull. 61:61-
109.
Bertelsen, E. 1951. The ceratioid fishes. Ontogeny, taxonomy, dis-
tribution and biology. Dana-Rep. Carlsberg Found. 39.
. 1958. The argenlinoid fish Xenoplhalmichthys danae. Dana-
Rep. Carlsberg Found. 45.
. 1980. Notes on Linophrynids. V. A revision of the deep-sea
anglerfishes (Ceratioidci) of the Lituyphryne arborifera-group.
Steenstrupia 6(6):29-70.
LITERATURE CITED
693
— . 1982. Notes on Linophnnidae. VIII. A review of the genus
Lmophryne. with new records and descriptions of two new species.
Steenstrupia 8(3):49-104.
— . 1983. First records of metamorphosed males of the families
Diceratiidae and Centrophrvnidae. Steenstrupia 8( 16):309-316.
— , G. Krefft and N. B. M,'\rsh.\ll. 1976. The fishes of the
family Notosudidae. Dana-Rep. Carlsberg Found. 86.
— , .AND N. B. Marshall. 1956. The Mirapinnati. a new order
of teleost fishes. Dana-Rep. Carlsberg Found. 42.
— , AND . 1958. Notes on Minpinnati (an addendum to
Dana-Rep. 42). Dana-Rep. Carlsberg Found. 45:9-10.
— , AND O. Mlink. 1964. Rectal light organs in the argentinoid
fishes Opislhoproctus and IVinleria. Dana-Rep. Carlsberg Found.
62.
— , AND T. W. PiETSCH. 1975. Results of research cruises of FRV
"Walther Herwig" to South Amenca. XXXVIIl. Osteology and
relationships of the ceratioid anglerfish genus Spiniphryne (family
Oneirodidae). .Arch. Fischereiwiss. 26(1): 1-1 1.
— , AND R. J. Lavenberg. 1981. Ceratioid anglerfishes of
the family Giganlactinidae: morphology, systematics, and distn-
bution. Nat. Hist. Mus. Los Ang. Cty. Contnb. Sci. 332.
— . AND P. J. Struhsaker. 1977. The ceratioid fishes of the genus
Thaumalichlhys: osteology, relationships, distnbution and biology.
Galathea Rep. 14:7-40.
-. B. Theisen AND O. MuNK. 1965. On a post-larval specimen.
anal light organ, and tubular eyes of the argentinoid fish Rhyn-
chohyalus natalensis (Gilchnst and von Bonde). Vidensk. Medd.
Dan. Naturhist. Foren. 128:357-371.
Bertin, L. 1937. Les poissons abyssau.x du genre Cycma Gunther
(anatomie, embryologie, bionomie). Dana-Rep. Carlsberg Found.
10:1-30.
Bertmar, G. 1959. On the ontogeny of the chondral skull in Charac-
idae, with a discussion on the chondrocranial base and the visceral
chondrocranium in fishes. Acta Zool. (Stockh.) 40:203-364.
Bertolini, F. 1933a. Famiglia I: Apogonidae, p. 306-309. In: Uova,
larve e sladi giovanili di Teleostei. Fauna Rora Golfo Napoli
Monogr. 38.
. 1933b. Famiglia 2: Serranidae. p. 310-331. In: Uova. larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
Beyer, J. E., AND G. C. Laurence. 1981. Aspects of stochasticity in
modelling growth and survival of clupeoid fish larvae. Rapp. P. V.
Reun. Cons. Int. Explor. Mer 178:17-23.
Bhargava, H. N. 1958. The development of the chondrocranium of
Maslacemhelus armalus (Cuv. et Val.). J. Morphol. 102:401-426.
Bhimachar, B. S., and A. David. 1945. On the occurrence of "external
gills" in the \oach — Lepidocephalus ihermalis (C. V.). Curr. Sci.
14(3):74,
BiGELOW, H. B., AND W. C. Schroeder. 1953. Fishes of the Gulf of
Maine. U.S. Fish Wildl. Serv. Fish. Bull. 53.
, AND . 1963. Family Osmeridae, p. 553-597. /«■ Fishes
of the western North Atlantic. H. B. Bigelow (ed.). Mem. Sears
Found. Mar. Res. No. 1 (Pt. 3).
, AND W. W. Welsh. 1925. Fishes of the Gulf of Maine. Bull.
Bur. Fish. Wash. 40:1-567.
BiNi, G. 1971. Atlante dei pesci delle Coste italiane. Vol. 3. Mondo
Sommerso, Milano.
BiRDSONG, R. S. 1975. The osteology of Microgobius signalus Poey
(Pisces: Gobiidae), with comments on other gobiid fishes. Bull. Fl.
State Mus. Biol. Sci. 19(3): 135-1 87.
Blaber.S. J., D. P. CiRL SAND A. K.Whitfield. 1981. The influence
of zooplankton food resources on the morphology of the estuanne
clupeid Gikhnstella aestuanus (Gilchnst. 1914). Environ. Biol.
Fishes 6:351-355.
Blache, J. 1964. Sur la presence de Luvarus impenales Raf 1810
dans I'Atlantique onental sud revelee par la decouverte de deux
larves au Stade Hystncinella de L. Roule (1924) (Pisces, Teleostei,
Perciformi. Luvareidei, Luvariidae) Cah. ORSTOM Ser. Oceanogr.
5:57-59.
. 1977. Leptocephales des poissons Anguilliformes dans la zone
sud du Golfe de Guinee. Faune Trop. 20:1-381.
Blackburn, J. E. 1973. A survey of the abundance and distribution
of ichthyoplankton in Skagit Bay, Unpubl. MS thesis. Univ. Wash-
ington, Seattle.
Blackburn, M. 1950. The Tasmanian whitebait, Loveltia seali (io\\i\-
ston) and the whitebait fishery. Aust. J. Mar. Freshwater Res.
1:155-198.
Blackler, R. W. 1974. Recent advances in otolith studies, p. 67-90.
In: Sea fisheries research. F. R. Harden Jones (ed.). John Wiley &
Sons, New York.
Blaxter, J. H. S. 1965. The feeding of herring larvae and their ecology
in relation to feeding. Calif. Coop. Oceanic Fish. Invest. Rep. 10:
78-88.
. 1966. The effect of light intensity on the feeding ecology of
herring, p. 393-409. In: Light as an ecological factor. R. Bainbndge.
G. C. Evans and O. Rackman (eds). Blackwell, Oxford.
. 1968. Rearing herring larvae to metamorphosis and beyond.
J. Mar. Biol. Assoc. U.K. 43:17-28.
. 1969a. Development: eggs and larvae, p. 177-252. In: Fish
physiology. Vol. 3. W. S. Hoar and D. J. Randall (eds.). Academic
Press, New York.
. 1969b. Expenmental reanng of pilchard larvae, Sardina pil-
chardus. J. Mar. Biol. Assoc. U.K. 49:557-575.
. 1971. Feeding and condition of Clyde herring larvae. Rapp.
P. V. Reun. Cons. Int. Explor. Mer 160:129-136.
. (ED.). 1974. The early life history of fish. Springer- Verlag,
Berlin.
. 1976. Reared and wild fish — how do they compare? Tenth
Europ. Symp. Mar. Biol., Ostend, Belgium. Sept. 17-23. 1975, Vol.
1:11-26.
, J. A. Gray AND A. C. Best. 1983. Structure and development
of the free neuromasts and lateral line system of the herring. J, Mar.
Biol. Assoc. U.K. 63:247-260.
, AND G. Hempel. 1963. The influence of egg size on herring
larvae. J. Cons. Int. Explor. Mer 28:21 1-240.
, AND F. G. T. HoLLiDAY. 1963. The behaviour and physiology
of herring and other clupeids. Adv. Mar. Biol. 1:261-393.
, AND J. R. Hunter. 1982. The biology of the clupeoid fishes.
Adv. Mar. Biol. 20:1-223.
, AND M. E. Staines. 1971. Food searching potential in manne
fish larvae, p. 467-485. In: Fourth European manne biology sym-
posium. D. J. Crisp (ed.). University Press, Cambndge.
Bleeker, P. 1854-1857. Bidrage tot de kennis der Sphyraenoiden van
den Indischen Archipel. Verh. Baravi Genoot. 26:1-22.
. 1 874. Equisse d'un systeme naturel des Gobioides. Arch. Neerl.
Sci. 9:289-331.
Block, B. A. 1983. Brain heaters in the billfish. Am. Zool. 23(4):936.
(Abstract.)
Bock, W. J. 1969. The concept of homology. Ann. N.Y. Acad. Sci.
167:71-73.
BoHLKE, J. E. 1957. A review of the blenny genus Chaenopsis. and
the description of a related new genus from the Bahamas. Proc.
Acad. Nat. Sci. Phila. 109:81-122.
. 1960a. Comments on serranoid fishes with disjunct lateral
lines, with the descnption of a new one from the Bahamas. Not.
Nat. (Phila.) 330:1-1 1.
. 1960b. Two new Bahaman species of the clinid fish genus
Paraclinus. Not. Nat. (Phila.) 337:1-8.
. 1 966. Order Lyomeri. Deep-sea gulpers, p. 603-628. In: Fish-
es of the western North Atlantic. G. W. Mead (ed.). Mem. Sears
Found. Mar. Res. No. l(Pt. 5).
. 1978. Muraenidae, Ophichthidae. In: FAO species identifi-
cation sheets for fishery purposes. Western Central Atlantic, Fishing
Area 31. Vol. 3. W. Fischer (ed.). FAO, Rome.
— , AND D. K. Caldwell. 1961. On the occurrence of the dac-
tyloscopid fish genus Daclylagnus in the west Atlantic. Bull. Mar.
Sci. Gulf Caribb. 11:537-542.
694
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
, AND C. C. G. Chaplin. 1968. Fishes of the Bahamas and
adjacent tropical waters. Livingston Publ. Co.. Wynnewood, PA.
, AND C. R. Robins. 1 959. Studies on fishes of the family Ophi-
diidae. II. Three new species from the Bahamas. Proc. Acad. Nat.
Sci. Phila. 111:37-52.
, and C. H. Robins. 1974. Description of a new genus and
species of clinid fish from the western Caribbean, with comments
on the families of the Blennioidea. Proc. Acad. Nat. Sci. Phila.
126(1): 1-8.
, AND D. G. Smith. 1968. A new xenocongrid eel from the
Bahamas, with notes on other species of the family. Proc. Acad.
Nat. Sci. Phila. 120:25-43.
, AND V. G. Springer. 1961. A review of the Atlantic species
of the clinid fish genus Slarksia. Proc. Acad. Nat. Sci. Phila. 1 13(3):
29-60.
,AND . 1975. A new genus and species offish (A'e/>!ac//>7i«
aleleslos) from the western Atlantic (Perciformes: Clinidae). Proc.
Acad. Nat. Sci. Phila. 127(7):57-61.
BOLIN, R. 1930. Embryonic development of the labrid fish 0.vi7w//i
ca/(/ormcM5 Giinther. Copeia 1930(4):I22-I28.
. 1934. Studies on California Cottidae: an analysis of the prin-
ciples of systematic ichthyology. Unpubl. PhD dissertation. Stan-
ford Univ., Stanford, CA.
. 1936a. Embryonic and early larval stages of the California
anchovy. Calif Fish Game 22:314-321.
. 1936b. The systematic position of Indoslomus paradoxus
Prashad and Mukerji, a new fresh water fish from Burma. J. Wash.
Acad. Sci. 26:420-423.
. 1939a. A new stomiatoid fish from California. Copeia 1939(1):
39-41.
. 1939b. A review of the myctophid fishes of the Pacific Coast
of the United States and of lower California. Stanford Ichthyol.
Bull. 1(4):89-156.
. 1941. Embryonic and early larval stages of the cottid fish
Orthonopias Iriacis Starks and Mann. Stanford Ichthyol. Bull. 2(3):
73-82.
Bond, C. E., and T. Uyeno. 1981. Remarkable changes in the ver-
tebrae of perciform fish Scombrolabrax with notes on its anatomy
and systematics. Jpn. J. Ichthyol. 28:259-262.
BoNHAM, K., and W. H. Bavliff. 1953. Radiography of small fishes
for meristic studies. Copeia 1953:150-151.
Booth, R. A. 1967. A description of the larval stages of the tomcod,
Microgadus tomcod. with comments on its spawning ecology. Un-
publ. PhD dissertation. Univ. Connecticut, Storrs.
Bortone, S. a. 1977. Osteological notes on the genus Centroprislis
(Pisces: Serranidae). Northeast Gulf Sci. 1:23-33.
Boulenger, G. a. 1902. Notes on the classification of teleostean
fishes. IV. On the systematic position of the Pleuronectidae. Ann.
Mag. Nat. Hist., Ser. 7. 10:295-304.
. 1904. Teleostei. Cambndge Nat. Hist. (Lond.) 7:539-727.
. 1915. Catalogueofthefreshwaterfishesof Africa in the British
Museum (Natural History). Vol. 3.
Bowman, A. 1920. The eggs and larvae of the angler (Lophius pis-
calorius L.) in Scottish waters. Sci. Invest. Fish. Board Scotl. 2:1-
42.
Boyd, J. F., and R. C. SiMMONDS. 1974. Continuous laboratory pro-
duction of fertile Fundulus heieroditus (Walbaum) eggs lacking
chorionic fibrils. J. Fish Biol. 6:389-394.
BoYDE, A., AND C. Wood. 1969. Preparation of animal tissues for
surface scanning electron microscopy. J. Microsc. 90:221.
Bradbury, M. G. 1967. The genera of batfishes (family Ogcocephal-
idae). Copeia 1967:399-422.
Branson, B. A., and G. A. Moore. 1962. The lateralis components
of the acoustico-lateralis system in the sunfish family Centrarchi-
dae. Copeia 1962:1-108.
Brauer, a. 1902. ijber einige von der Valdivia-Expedition gesam-
melte Tiefseefische und ihre Augen. Sitzungsber. Ges. Beford. Na-
turwiss. Marburg (1901) nr. 8:115-130.
. 1906. DieTiefsee-Fische. I.SystematischerTeil. Wiss. Ergebn.
Dtsch. Tiefsee-Exped. Valdivia 15(1).
. 1908. DieTiefsee-Fische. II. AnatomischerTeil. Wiss. Ergebn.
Dtsch. Tiefsee-Exped. Valdivia 15(2).
Bray, D. 1983. A revision of the family Zeidae. Unpubl. MSc thesis.
Australia. (Complete reference not available.)
Breder, C. M., Jr. 1923. Some embryonic and larval stages of the
winter flounder. Bull. Bur. Fish. Wash. 38:31 1-315.
. 1927. Scientific results of the first oceanographic expedition
of the "Pawnee" 1925. Fishes. Bull. Bingham Oceanogr. Collect.
Yale Univ. l(l):l-90.
. 1932. On the habits and development of certain Atlantic
Synentognathi. Carnegie Inst. Wash. Publ. 435, Pap. Tortugas Lab.
28(l):l-35.
. 1938. A contribution to the life histories of Atlantic flying
fishes. Bull. Bingham Oceanogr. Collect. Yale Univ. 6(5):1-121.
. 1939. On the life history and development of the sponge
blenny, Paraclmus marmoratus (Steindachner). Zoologica (N.Y.)
24:487-496.
. 1941. On the reproductive behavior of the sponge blenny,
Paraclinus marmoratus (Steindachner). Zoologica (N.Y.) 26:233-
236.
. 1 944. Materials for study of the life history of Tarpon atlan-
ticus. Zoologica (N.Y.) 29:217-252.
. 1959, Observations on the spawning behavior and egg de-
velopment of Strongylura notaia (Poey). Zoologica (N.Y.) 44: 141-
147.
. 1962. On the significance of transparency in osteichthid fish
eggs and larvae. Copeia 1962:561-567.
, and E. Clark. 1947. A contribution to the visceral anatomy,
development, and relationships of the Plectognathi. Bull. Am. Mus.
Nat. Hist. 88:287-320.
, and P. Rasquin. 1952. The sloughing of the melanic area of
the dorsal fin, an ontogenic process in Tylosurus raphidoma. Bull.
Am. Mus. Nat. Hist. 99(1): 1-24.
, AND . 1954. The nature of post-larval transformation
in Tylosurus acus (Lacepede). Zoologica (N.Y.) 39: 1 7-30.
, AND D. E. Rosen. 1966. Modes of reproduction in fishes. T.
F. H. Publ., Inc., Neptune City, New Jersey.
Brembach, M. 1976. Anatomische Beitrage zur Systematik lebendge-
barender Halbschnabler (Hemirhamphidae, Pisces). Z. Zool. Syst.
Evolutionsforsch. 14(3): 169-1 77.
Briggs, J.C. 1952. Systematic notes on the oceanic fishes of the genus
Lophotus. Copeia 1952:206-207.
. 1955. A monograph of the clingfishes (order Xenopterygii).
Stanford Ichthyol. Bull. 6:1-224.
Brill, J. 1981. Cremchthys baileyi (GxXhen. 1983). J. Am. Killifish
Assoc. 14(5):172-184.
Brinley, F. J. 1939. Spawning habits and development of beau-greg-
ory (Pomacenlrus leucosliclus). Copeia 1939:185-188.
Brook, G. 1890. Notes on larval stages of Afotella. Proc. R. Phys.
Soc. Edinb. 10:156-161.
Brothers, E. B. 1981. What can otolith microstructure tell us about
daily and subdaily events in the early life history of fish? Rapp.
P.V. Reun. Cons. Int. Explor. Mer 178:393-394.
, C. P. Mathews and R. Lasker. 1976. Daily growth incre-
ments in otoliths from larval and adult fishes. Fish. Bull. U.S. 74:
1-8.
, and W. N. McFarland. 1981. Correlations between otolith
microstructure, growth, and life history transitions in newly re-
cruited French grunts, [ffaemulon /lavolincalum (Desmarest), Hae-
mulidae]. Rapp. P.V. Reun. Cons. Int. Explor. Mer 178:369-374.
— , E. D. Prince AND D. W. Lee. 1984. Age and growth of young-
of-the-year bluefin tuna (Thunmis thynnus) from otolith micro-
structure. NOAA Tech. Rep. 8. In press.
, AND R. E. Thresher. MS. Interspecific differences in the du-
ration of the pelagic larval stage of Western Atlantic angelfishes
(Pomacanthidae): possible biogeographic and systematic applica-
tions.
LITERATURE CITED
695
, D. M. Williams AND P. F. Sale. 1983. Length of larval life
in 12 families of fishes at One Tree Lagoon. Great Barrier Reef,
Australia. Mar. Biol. (Berl.) 76:319-324.
Brownell, C. L. 1979. Stages in the early development of 40 marine
fish species with pelagic eggs from the Cape of Good Hope. Rhodes
Univ. J. L. B. Smith Inst. Ichthyol. Bull. 40.
. 1980a. Water quality requirements for first-feeding in marine
fish larvae. 1. Ammonia, nitrite, and nitrate. J. Exp. Mar. Biol. Ecol.
44:269-283.
1 980b. Water quality requirements for first-feeding in marine
fish larvae. II. pH, oxygen, and carbon dioxide. J. Exp. Mar. Biol.
Ecol. 44:285-298.
Bruggeman. E., P. Sorgeloos AND P. Van Haecke. 1980. Improve-
ments in the decapsulation techniqueof.-Jr/e?«;a cysts, p. 261-269.
In: The brine shrimp Artemia. Vol. 3. Ecology, culturing, use in
aquaculture. G. Persoone, P. Sorgeloos, O. Roels, and E. Jaspers
(eds.). Universa Press, Wetteren, Belgium.
Brummett. a. R., and J. N. Dumont. 1981. A comparison of cho-
rions from eggs from northern and southern populations of Fun-
dulus helerocUtus. Copeia 1981:607-614,
Brundin, L. 1966. Transantarctic relationships and their significance
as evidenced by chironomid midges. K. Sven. Vetenskapsakad.
Handl. 4(11)1-472.
Bri'Ton, M, N. 1979. The breeding biology and early development
of Clanas gariepinus (Pisces: Clariidae) in Lake Sibaya, South Af-
rica, with a review of breeding in species of the subgenus Clarias
(Clarias). Trans. Zool. Soc. Lond. 35:1-45.
Bruun, A. F. 1935. Flying-fishes (Exocoetidae) of the Atlantic. Dana-
Rap. Carlsberg Found. 6.
. 1937. Chascanopselta in the Atlantic; a bathypelagic occur-
rence of flatfish with remarks on distribution and development of
certain other forms. Vidensk. Medd. Dan. Naturhist. Foren. 101:
125-135.
. 1 940. A study of a collection of the fish Schindlena from south
Pacific waters. Dana-Rep. Carlsberg Found. 21.
Bryan, P. G., and B. B. Madraisall 1977. Larval rearing and de-
velopment of Siganus Imeatus (Pisces: Siganidae) from hatching
through metamorphosis. Aquaculture 10:243-252.
BiDD, P. L. 1940. Development of the eggs and early larvae of six
California fishes. Calif Dep. Fish Game Fish. Bull. 56.
Budgett, J. S. 1902. The habits and development of some West
African fishes. Proc. Camb. Philos. Soc. 1 1:102-104.
. 1903. On the breeding-habits of some West-African fishes.
with an account of the external features in development o{ Prolop-
lerus annectens. and a description of the larva of Prowpterus la-
pradei. Trans. Zool. Soc. Lond. 16:1 15-136.
Burdak, V. D. 1969. Ontogenesis of squamation in the mullet (Mugil
saliens Risso). Zool. Zh. 48:241-247.
Burgess. W. E. 1965. Larvae of the surgeonfish genus .-lfa«//n«-«i of
the tropical western Atlantic. Unpubl. MS thesis. Univ. Miami,
Coral Gables, FL.
. 1974. Evidence for the elevation to family status of the an-
gelfishes (Pomacanthidae), previously considered to be a subfamily
of the butterfly-fish family, Chaetodontidae. Pac. Sci. 28:57-71.
. 1978. Butterflyfishes of the worid. TFH Publications, Ltd.
Neptune City, NJ.
. 1982. The first aquarium spawning of the woodcat, Trachy-
corystes insignis. Trop. Fish Hobby. 1982 (August):84-89.
Burke, V. 1930. Revision of the fishes of the family Liparidae. U.S.
Natl. Mus. Bull. 150:1-204.
Bussing, W. A. 1965. Studiesof the midwater fishes of the Peru-Chile
Trench. Antarct. Res. Ser. 5:185-227.
Bitler, J. L. 1979. The nomeid genus Cubiceps (Pisces) with a de-
scription of a new species. Bull. Mar. Sci. 29:226-241.
, and E. H. Ahl-STROM. 1976. Review of the deep-sea fish genus
Scopetengys (Neoscopelidae) with a description of a new species.
Scopelengys clarkei, from the central Pacific. Fish. Bull. U.S. 74:
142-150.
, H. G. Moser, G. S. Hageman and L. L. Nordgren. 1982.
Developmental stages of three California sea basses (Paralabrax.
Pisces, Serranidae). Calif Coop. Oceanic Fish. Invest. Rep. 23:252-
268.
Byrne, D. M. 1978. Life history of the spotfin killifish. Fundutus luciae
(Pisces: Cypnnodontidae), in Fox Creek Marsh, Virginia. Estuaries
1:211-227.
Cadwallader, p. L. 1976. Breeding biology of a non-diadromous
galaxiid Galaxias vulgaris Stokell in a New Zealand river. J. Fish
Biol. 8:157-177.
Cailliet, G. M., and R. N. Lea. 1977. Abundance of the "rare"
zoarcid, Maynea californica Gilbert. 1915, in the Monterey Can-
yon. Monterey Bay, California. Calif Fish Game 63:253-261.
Caldwell, D. K. 1962. Development and distribution of the short
bigeye, Pseudopriacanlhus alius (Gill) in the western North Atlan-
tic. U.S. Fish Wildl. Serv. Fish. Bull. 62(203): 103-150.
Caldwell, M. C. 1962. Development and distnbution of larval and
juvenile fishes of the family Mullidae of the western North Atlantic.
U.S. Fish Wildl. Serv. Fish. Bull. 62(213):403-457.
Callegrini, C, and F. Basaglia. 1978. Biochemical characteristics
of Mugilids in the lagoons of the Po Delta. Boll. Zool. 45(1):35-
40.
Cameron, J. 1959. The larval and post-larval stages of Gymnam-
modytes semisquamatus (Jourdain). J. Mar. Biol. Assoc. U.K. 38:
17-25.
Campana, S. E. 1983. Calcium deposition and otolith check formation
during periods of stress in coho salmon. Oncorhynchus kisutch.
Comp. Biochem. Physiol. 75A:2 15-220.
Campos, H. 1 969. Reproduccion del Aplochiton laematus Jenyns. Bol.
Mus. Nac. Hist. Nat. Chile 29:207-222.
. 1972. Breeding season and early development of i?rac/n'^a-
laxias bullock: (Osteichthyes: Galaxiidae). Tex. J. Sci. 23:53 1-544.
. 1979. Multivariate analysis of the taxonomy of the fish family
Galaxiidae. Zool. Anz. 202:280-288.
Cantwell, G. E. 1964. A revision of the genus Parapercis. family
Mugiloididae. Pac. Sci. 18:239-280.
Carbine. W. F. 1 944. Egg production of the northern pike. Esox lucius
L., and the percentage survival of eggs and young on the spawning
grounds. Pap. Mich. Acad, of Sci. Arts Lett. 29:123-137.
Cardeilhac, p. T. 1976. Induced maturation and development of
pinfish eggs. Aquaculture 8:389-393.
Carlstrom, D. 1963. A crystallographic study of vertebrate otoliths.
Biol. Bull. (Woods Hole) 125:441-463.
Carr, J. 1982. Aphyosemion ainieli Radda. 1976. J. Am. Killifish
Assoc. 15(3): 102- 104.
Carr, W. E. S., and C. A. Adams. 1972. Food habits of juvenile
marine fishes: evidence of the cleaning habit in the leatherjacket,
Oligoplites saunis. and the spottail pinfish, Diplodus holbronki. Fish
Bull. U.S. 70:1111-1120.
Caruso, J. H. 1981. The systematics and distribution of the lophiid
anglerfishes: I. A revision of the genus Lophioides with the descrip-
tion of two new species. Copeia 1981:522-549.
. 1983. The systematics and distribution of the lophiid angler-
fishes: II. Revisions of the genera Lophiomus and Lophius. Copeia
1983:11-30.
, AND H. R. BiiLLis, Jr. 1976. A review of the lophiid angler
fish genus Sladenia with a description of a new species from the
Caribbean Sea. Bull. Mar. Sci. 26:59-64.
Cassel, a. 1 98 1 . Garmanella pulchra. a fascinating killifish from Mex-
ico. J. Am. Killifish Assoc. 1 4(5): 1 67- 1 68.
Casteel, R. W. 1974. Identification of the species of Pacific salmon
(Genus Oncorhynchus) native to North .America based on otoliths.
Copeia 1974:305-311.
Castle, P. H. J. 1968. Larvaldevelopment ofthecongrid eel G«ar/io-
phis capensis (Kaup), ofl^ southern Africa, with notes on the identity
of Congermuraena auslralis Barnard. Zool. Afr. 3:139-154.
. 1969. An index and bibliography of eel larvae. Rhodes Univ.
J. L. B. Smith Inst. Ichthyol. Spec. Publ. 7:1-121.
. 1973. A giant notacanthiform leptocephalus from the Chat-
696
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
ham Islands, New Zealand. Rec. Dom. Mus. Wellington 8(8): 121-
124.
. 1976. Criteria of satisfactory fixation and preservation of eel
larvae (Leptocephali), p. 319-321. In: Zooplankton fixation and
preservation. H. F. Steedman (ed.). Unesco Press, Pans.
. 1979. Early life-history of the eel Moringua edwardsi (Pisces,
Moringuidae) in the western North Atlantic. Bull. Mar. Sci. 29:1-
18.
. 1980. Larvae of the ophichthid eel genus Necnchelys in the
Indo-Pacific. Pac. Sci. 34:165-171.
, .AND J. E. BoHLKE. 1976. Sexual dimorphism in sizc and ver-
tebral number in the western Atlantic eel Moringua edwardsi (An-
guilliformes, Moringuidae). Bull. Mar. Sci. 26:615-619.
, AND N. S. Raju. 1975. Some rare leptocephali from the At-
lantic and Indo-Pacific oceans. Dana-Rep. Carlsberg Found. 85:1-
25.
, AND D. A. Robertson. 1974. Early life history of the congrid
eels Gnalhophis hahenatus and G. incogniliis in New Zealand waters.
N.Z. J. Mar. Freshwater Res. 8:95-1 10.
Cavender,T. M. 1969. An Oligocenemudminnow (family Umbridae)
from Oregon with remarks on relationships within the Esocoidei.
Occas. Pap. Mus. Zool. Univ. Mich. No. 660.
. 1970. A comparison of coregonines and other salmonids with
the earliest known teleost fishes, p. 1-32. In: Biology of coregonid
fishes. C. C. Lindsey and C. S. Woods (eds.). Univ. Manitoba Press,
Winnipeg.
Cerny, K. 1977. The early development of chub LeMmcM.5 cep/ia/iu
(L., n5S),rudd-Scardiniuserythrophthalmus{L., 1758)and roach-
RutilusruulusCL., 1758). Acta Univ. Carol. Biol. 1977(1-2): 1-149.
Cervigon M., F. 1966. Los peces marinos de Venezuela. Fund. La
Salle Cienc. Nat. (Caracas), Vol. I and 2.
, and E. Velazquez. 1978. Revision systematica y estudio
osteologico de las especies del genero Jenkinsia de las costas de
Venezuela (Osteichthyes: Clupeiformes). Cuad. Oceanogr. Univ.
Onente 7:3-54.
Chabanaud, p. 1934. Contribution a la morphologic des poissons
heterosomes. Bull. Soc. Zool. Fr. 59:123-129.
. 1937. L'extension prorse peripherique, la contraction axiale
post-hypophysaire et I'anisoconie rhachidienne des teleosteens dys-
symetriques. Bull. Soc. Zool. Fr. 62:368-385.
. 1940. Notules ichtyologiques. Bull. Mus. Natl. Hist. Nat. (2)
12:149-156.
. 1941. Sur la classification et la geonemie de soleides du genre
— . 1948. The osteology and relationships of the Microstomidae,
a family of oceanic fishes. Proc. Calif Acad. Sci., Ser. 4, 26:1-22.
— , andL. D. TowNSEND. 1938. 1\te o^leoXogy of Zaprora silenus.
Aesopia. Copeia 1941:31-32.
. 1949. Le probleme de la phylogenese des Heterosomata. Bull.
Inst. Oceanogr. (Monaco) 950:1-24.
. 1950. Contribution a I'anatomie et a la systematique de la
famine des Bothidae, s. str. Bull. Soc. Zool. Fr. 74:246-253.
. 1969. Sur la morphologie de Samans cristalus Gray et de
Samaris cristalus erythraeus P. Chabanaud natio nova (Pleuronec-
toidea Samaridae). Bull. Mus. Natl. Hist. Nat. (2) 40:874-890.
Chacko, p. I. 1944. On the bionomics of the big-jawed jumper, Lac-
tarius lactarius (Cuv. and Val.). Curr. Sci. (Bangalore) 13(4): 108.
Chambers, J. R., J. A. MusicK AND J. Davis. 1976. Methods of dis-
tinguishing larval alewife from larval blueback herring. Chesapeake
Sci. 17:93-100.
Chang, H., X.Sha,Z. Chen, H. Rl an and G. He. 1981. Observations
on the egg and larval development of a clupeid Harengula zunasi
Bleeker. Trans. Chin. Ichthyol. Soc. 1:57-64.
, , G. He AND L. Song. 1980. A description of the mor-
phological characters of the eggs and larvae of the flathcad fish,
Ptatycephalus indicus. Oceanol. Limnol. Sin. 1 1(2):I61-168.
Changjiang, Y. 1976. Fishes of the Yangtze River. Hunan Province
Inst. Hydrobiol., Science Press, Peking. (In Chinese.)
Chapman, W. M. 1939. Eleven new species and three new genera of
oceanic fishes collected by the International Fisheries Commission
from the northeastern Pacific. Proc. U.S. Natl. Mus. 86:501-542.
. 1942. The osteology and relationships of the Argentinidae, a
family of oceanic fishes. J. Wash. Acad. Sci. 32:104-1 17.
with notes on its distribution and early life history. Ann. Mag. Nat.
Hist., Ser. 11, 2:89-117.
Chaudhuri, H. 1962. Induced breeding and development of a com-
mon catfish, Ornpok bimaculatus (Bloch). Proc. Indian Sci. Congr.
49:390-391.
, J. V. JuARio, J. H. Primavera, R. Samson and R. Mateo.
1978. Observations on artificial fertilization of eggs and the em-
bryonic and larval development of milkfish, Chanos chanos (For-
skal). Aquaculture 13:95-1 13.
Chen, F. Y., M. Chow, T. M. Cha^) and R. Lim. 1977. Artificial
spawning and larval rearing of the grouper, Epinephelus tauvina
(Forskal) in Singapore. Singapore J. Primary Ind. 5:1-21.
Chenoweth, S. B. 1970. Seasonal variations in condition of larval
herring in Boothbay Harbor area of the Maine coast. J. Fish. Res.
Board Can. 27:1875-1879.
Chereshnev, I. A., AND A. V. Balushkin. 1980. A new species of
blackfish, Dallia asmirabilis sp. n. (Umbridae, Esociformes) from
the Amguema River basin (Arctic Chukotka). J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 20(6):25-30.
Chernyayev, Zh. a. 1971. Some data on the reproduction and de-
velopment of the cottoid fish (Comephorus dyhowskii Korotneff).
J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 1 1:706-716.
. 1975. Morphological and ecological features of the reproduc-
tion and development of the "Big Golomyanka" or Baikal oil-fish
(Comephorus baicalensis). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
14:856-868.
. 1978. Morpho-ecological characteristics of the reproduction
and development of the Baikal sand sculpin, Paracoltus {Leocoltus)
kessleri. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 17:890-903.
. 1981. Reproduction and development ofthe "Big-head" scul-
pin, Balrachocollus baicalensis. from Lake Baikal. J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) l9(6):83-94.
Chidambaram, K. 1942. Observations on the development of Anus
jella (Day). Proc. Indian Acad. Sci. Sect. B 14:502-508.
Chkhrinskiv, a. I. 1970. The nature of oogenesis and fecundity in
the Japanese horsemackerel [Trachurus japonicus (Temminck et
Schlegel)]. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 10:755-760.
Chirichigno, N. 1973. Nuevas especies de peces de los generos A/ui-
lelus (fam. Triakidae), Raja (fam. Rajidae) y Schedophilus (fam.
Centrolophidae). Inst. Mar. Peru (Callao) Inf 42.
Chrapkova-Kovalevskaya, N. V. 1963. Data on reproduction, de-
velopment and distribution of larvae and young fish oi Oxyporhani-
phus micropterus Val. (Pisces, Oxyporhamphidae). Tr. Inst. Okean-
ol. Acad. Nauk SSSR 62:49-61.
Chung, M. 1961. Illustrated encyclopedia. The fauna of Korea (2).
Fishes.
. 1977. The fishes of Korea. II Ji Sa Publishing Co., Seoul.
CiPRiA. G. 1927. Contributo alia conoscenza dello sviluppo postem-
brionale di Hypsirhynchus hepalicus Facciola. Mem. R. Com. Tal-
assogr. Ital. 130.
. 1934. Uova, stadi embrionali e post-embrionali di Blennius
palmicornis C\i\ . Mem. R. Com. Talassogr. Ital. 218.
. 1936. Uova, stadi embrionali e post -embnonali, nei Blennidi.
I. Blennius pavo Risso. II. Blennius maequalis C. V. Mem. R. Com.
Talassogr. Ital. 231.
Clancey, J. F. 1956. A contribution to the life history of the fish,
Bregmaceros atlanticus Goode and Bean, from the Florida Current.
Bull. Mar. Sci. Gulf Camb. 6:233-260.
Clark, E. 1950. Notes on the behavior and morphology of some West
Indian plectognath fishes. Zoologica (N.Y.) 35:159-168.
Clark, F. N. 1925. The life history of Leurest ties tenuis, an atherine
fish with tide controlled spawning habits. Calif Dep. Fish Game
Fish Bull. 10:1-51.
. 1929. The life history of the California jack smelt, Atherinopsis
californiensis. Calif. Dep. Fish Game Fish Bull. 16:1-22.
Clark, J., W. G. Smith, A. W. Kendall, Jr. and M. P. Fahav. 1 969.
LITERATURE CITED
697
Studies of estuarine dependence of Atlantic coastal fishes. U.S. Bur.
Sport Fish. Wild!. Tech. Pap. 28.
Clark, R. S. 1914. General report on the larval and post-larval te-
leosteans in Plymouth waters. J. Mar. Biol. Assoc. U.K. 10:327.
. 1920. The pelagic young and early bottom stages of teleosteans.
J. Mar. Biol. Assoc. U.K. 12:159-240.
Clakke, T. a. 1973. Some aspects of the ecology of lantemfishes
(Myctophidae) in the Pacific Ocean near Hawaii. Fish. Bull. U.S.
71:401-434.
, AND P. J. Wagner. 1976. Vertical distribution and other
aspects of the ecology of certain mesopelagic fishes taken near Ha-
waii. Fish Bull. U.S. 74:635-645.
Clausen, H. S. 1959. Denticipitidae, a new family of isospondylous
teleosls from west African freshwater. Vidensk. Medd. Dan. Nat-
urhist. Foren. 121:141-156.
Cohen, D. M. 1958. A revision of the fishes of the subfamily Argen-
tininae. Bull. Fl. State Mus. Biol. Ser. 3:93-172.
. 1 960. New records of the opisthoproctid genus Balhylychnops.
with a notice of neoteny in the related genus DoUchopleryx. Copeia
1960:147-149.
. 1964a. Bioluminescence in the Gulf of Mexico anacanthine
fish Steindachnena argentea. Copeia 1964:406-409.
. 1964b. Suborder Argentmoidea. p. 1-70. In: Fishes of the
western North Atlantic. H. B. Bigelow (ed.). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 4).
. 1973. Eretmophoridae, p. 322-326. /«.■ Check-list of the fishes
of the north-eastern Atlantic and of the Mediterranean (Clofnam).
Vol. I. J. C. Hureau and T. Monod (eds.). UNESCO, Pans.
. 1974. A review of the pelagic ophidioid fish genus Brotula-
taenia with descriptions of two new species. Zool. J. Linn. Soc. 55:
119-149.
. 1979. Notes on the mond fish genera L()?e//a and P/!.vi;cM/j«
in Japanese waters. Jpn. J. Ichthyol. 26:225-230.
, ANn J. G. Nielsen. 1978. Guide to the identification of genera
of the fish order Ophidiiformes with a tentative classification of the
order. NOAA Tech. Rep. NMFS Circ. 147.
, and J. L. Russo. 1979. Variations in the fourbeard rockling.
Enchelyopus cimhrius. a North Atlantic gadid fish, with comments
on the genera of rocklings. Fish. Bull. U.S. 77:91-104.
Cole, F. J., and J. Johnstone. 1902. Pleuronecles (the plaice). Mem.
Liverpool Mar. Biol. Comm. 8.
Coleman, L. R., and B. G. Nafpaktitis. 1972. Dorsadcna yaqiunae.
a new genus and species of myctophid fish from the eastern nonh
Pacific Ocean. Nat. Hist. Mus. Los Ang. Cty. Contrib. Sci. 225:1-
11.
Colin, P. L. 1975. The neon gobies. The comparative biology of the
gobies of the genus Gobwsoma. subgenus Elacatmus. (Pisces: Go-
biidae) in the tropical western north Atlantic Ocean. T.F.H. Pub-
lications, Neptune City, NJ.
. 1982. Spawning and larval development of the hogfish, Lflc/;-
nolaimus maximus. (Pisces, Labridae). Fish. Bull. U.S. 80:853-
862.
CoLLETT, R. 1905. Fiske indsamlede under "Michael Sars" togter i
nordhavet 1900-1902. Rep. Norw. Fish. Mar. Invest. 2(3):1-I47.
Coi LETTE, B. B. 1962. Hemiramphus bermudensis. a new halfbeak
from Bermuda, with a survey of endemism in Bermudian shore
fishes. Bull. Mar. Sci. Gulf Canbb. 12:432-449.
. 1966. Betomon. a new genus of fresh-water needlefishes from
South America. .Am. Mus. Novit. No. 2274.
. 1974a. South American freshwater needlefishes (Belonidae)
of the genus Pseudolylosunis. Zool. Meded. (Leiden) 48(16):I69-
186.
. 1974b. The garfishes (Hemiramphidae) of Australia and New
Zealand. Rec. Aust. Mus. 29(2): 1 1-105.
. 1974c. Sfrowgi'/wra /iififci. a new species of freshwater needle-
fish from the Usumacinta Province of Guatemala and Mexico.
Copeia 1974:611-619.
. 1979. Adaptations and systematics of the mackerels and tunas.
p. 7-39. In: The physiological ecology of tunas. G. Sharp and A.
Dizon (eds.). Academic Press, NY.
— . 1982a. South American freshwater needlefishes of the genus
Polamorrhaphis (BelomfOTmes: Belonidae). Proc. Biol. Soc. Wash.
95:714-747.
— . 1982b. Rediscovery of Hyporhamphus xanthoptenis. a half-
beak endemic to Vembanad Lake, Kerala, southern India. Matsya
7:1-12.
— . 1982c. Two new species of freshwater halfbeaks (Pisces: Hem-
iramphidae) of the genus Zenarchoplerus from New Guinea. Copeia
1982:265-276.
— . 1983. Recognition of two species of double-lined mackerels
(Grammatorcynus: Scombridae). Proc. Biol. Soc. Wash. 96:715-
718.
— , and F. H. Berry. 1965. Recent studies on the needlefishes
(Belonidae): An evaluation. Copeia 1965:386-392.
— , AND L. N. Chao. 1975. Systematics and morphology of the
bonilos (Sarda) and their relatives (Scombridae. Sardini). Fish.
Bull. U.S. 73:516-625.
— , andN. V. Parin. 1970. Needlefishes (Belonidae) of the eastern
Atlantic Ocean. Atl. Rep. 1 1:7-60.
— , AND J. L. Rlsso. 1979. An introduction to the Spanish mack-
erels, genus Scomberomonis. p. 3-16. In: Proc. Mackerel Collo-
quium. E. L. Nakamura and H. R. Bullis, Jr. (eds.). Gulf States
Mar. Fish. Comm.
, AND . In press. Morphology, systematics. and biology
of the Spanish mackeTeh(,Scomberomorus. Scombridae). Fish. Bull.
U.S.
Collins, L. A., J. H. Finucane and L. E. Barger. 1980. Description
of larval and juvenile red snapper, Lutjanus campechanus. Fish.
Bull. U.S. 77:965-974.
Collins, S. P., and M. P. Barron. 1981. Demersal and pelagic trawl-
ing survey of the M.T. "Denebola" in southern Australian waters,
1979-80 summer. Tasmanian Fish. Res. 24:2-48.
Colton, J. B., AND K. A. Honey. 1963. The eggs and larval stages of
the butterfish, Poronotus triacanlfius. Copeia 1963:447-450.
, AND R. R. Marak. 1969. Guide for identifying the common
planktonic fish eggs and larvae of continental shelf waters. Cape
Sable to Block Island. NMFS Woods Hole Biol. Lab. Ref. Doc.
69-9.
CoNAND, F. 1978, Systematique des larves de Clupeides de PAtlan-
tique oriental entre 20°N et 15°S (eaux marine et saumatres). Cah.
ORSTOM Ser. Oceanogr. 16:3-8.
, AND E. Fagetti. 1971. Description et distribution saisonniere
des larves de sardinelles des cotes du Senegal et de la Gamble en
1968 et 1969. Cah. ORSTOM Ser. Oceanogr. 9:293-318.
, AND C. Franqlieville. 1973. Identification et distribution
saisonniere de larves de Carangides au large du Senegal et de la
Gamble. Bull. Inst. Fondam. Afr. Noire, Ser. A Sci. Nat. 35:951-
978.
CoNSTANTZ, G. D. 1981. Life history patterns of desert fishes, p. 237-
290. In: Fishes in North American deserts. R. J. Naiman and D.
L. Soltz (eds.). John Wiley & Sons. N.Y.
Coombs, S. H., and A. R. Hiby. 1979. The development of the eggs
and early larvae of blue whiting, Micromesistius poiassou and the
effect of temperature on development. J. Fish Biol. 14:1 1 1-123.
Cooper, J. E. 1978. Identification of eggs, larvae, and juveniles of the
rainbow smelt, Osmerus morda.x. with comparisons to larval ale-
wife, Alosa pseudoharengus, and gizzard shad. Durosoma cepe-
dianum. Trans. Am. Fish. Soc. 107:56-62.
Costa, A. 1869. Osservacioni sul Krohnius filamentosus e sullo svi-
luppodella pinna caudale del Trachyptenis. .Ann. Mus. Zool. Napoli
5:1-41.
Courtenay, W. R., Jr. 1971. Sexual dimorphism of the sound pro-
ducing mechanism of the stnped cusk-eel, Rissola marginala (Pisces:
Ophidiidae). Copeia 1971:259-268.
, and F. A. McKiTTRicK. 1970. Sound-producing mechanisms
in carapid fishes, with notes on phylogenetic implications. Mar.
Biol. (Beri.) 7:131-137.
698
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Cressey, R. 1981. Revision of Indo-Pacific lizardfishes of the genus
Synodus (Pisces: Synodontidae). Smithson, Contrib. Zool. 342.
Cressey, R. P., B. B. CoLLETTE AND J. L. Rlisso. 1983. Copepodsand
scombrid fishes: A study in host-parasite relationships. Fish. Bull.
U.S. 81:227-265.
Crossland, J. 1981. Fish eggs and larvae of the Hauraki Gulf New
Zealand. N.Z. Minist. Agric. Fish. Res. Bull. 23.
• . 1982. Distribution and abundance of fish eggs and larvae from
the spring and summer plankton of north-east New Zealand, 1976-
78. N.Z. Minist. Agric. Fish. Res. Bull. 24.
Grossman, E. J. 1980. Esox amencanus. Esox lucius, Esox inasqui-
nongy. Esox niger. pp. 1 31-138. In: Atlas of North American fresh-
water fishes. D. S. Lee, C. R. Gilbert, C. H. Hocutt, R. E. Jenkins,
D. E. McAllister and Jay R. StaufTer, Jr. (eds.). North Carolina State
Mus. Nat. Hist., Raleigh, NC.
Cunningham, J. T. 1887. The eggs and larvae of teleosteans. Trans.
R. Soc. Edinb. 33:97-136.
. 1889. Studies of the reproduction and development of te-
leostean fishes occurring in the neighbourhood of Plymouth. J. Mar.
Biol. Assoc. U.K. 1:10-54.
. 1890. A treatise on the common sole {Solea vulgaris), con-
sidered both as an organism and as a commodity. Mar. Biol. Assoc.
U.K., Plymouth.
. 1891. On some larval stages of fishes. J. Mar. Biol. Assoc.
U.K. 2:68-74.
Curless, W. W. 1979. Early development and larval biology of the
California topsmelt, Atherinops affinis. Amer. Zool. 19:949. (Ab-
stract.)
CusHiNG, D. H. 1975. Marineecology and fisheries. Cambridge Univ.
Press, Cambridge.
. 1977. Science and the fisheries. Camelot Press Ltd., South-
ampton.
Cuvier,G. 1829. DerOreosome(Oreo50wa, nov.jetparticulierement
de rOreosome de I'Atlantique (Oreosoma atlanlicum. nov.), p. 515-
518. In: Histoire naturelledes poissons. 4. Des Acanthopterygiens.
G. Cuvier and A. Valenciennes (eds.). F. G. Levrault, Paris.
CzoLowsK.'V, R. 1962. Morphogenesis of skin scuta in S/p/!o«osfo/«a
typhle L. Zool. Pol. 12(3):269-303.
Daan, N. 1981. Comparison of estimates of egg production of the
Southern Bight cod stock from plankton surveys and market sta-
tistics. Rapp. P.V. Reun. Cons. Int. Explor. Mer 1 78:242-243.
Dakin, W. J., AND G, L. Kesteven. 1938. The Murray cod (Ma«-»-
lochella macquarensis (Cuv. et Val.)). Some experiments on breed-
ing with notes on the early stages and a reference to the problems
of depletion and restocking. State Fish. N.S.W. Res. Bull. I.
Dale, T. 1984. Embryogenesis and growth of otoliths in the cod (Gat/-
us morhua L.), p. 231-250. In: The propagation of cod. E. Dahl,
D. S. Danielssen, E. Moksness and P. Solemdal (eds.). Flodevigen
Rapp. 1.
D'Ancona, U. 1929. Note di ittiologia mediterranea. 1. Stadi, gio-
vanili e uova di Exocoetus. Pubbl. Staz. Zool. Napoli 9:203-21 1.
. 1931a. Clupeidae, p. 1-16. In: Uova, larve e stadi giovanili
di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1931b. Ordine: Apodes (Muraenoidei), p. 94-156. In: Uova.
larva e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli
Monogr. 38.
• . 1933a. Famiglia 1: Gadidae, p. 178-255. In: Uova, larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1933b. Famiglia: Lampridae. p. 280. In: Uova, larve e stadi
giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
— . 1933c. Famiglia: Syngnathidae, p. 281-298. In: Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
— . 1933d. Famiglia: Macrorhamphosidae, p. 299-306. /«.• Uova,
larve e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli
Monogr. 38.
— , AND G. Cavinato. 1965. The fishes of the family Bregma-
cerotidae. Dana-Rep. Carlsberg Found. 64.
Dando, p. R. 1975. On the northern rockling Ciliata septentrionalis
(Collett) in the Plymouth area. J. Mar. Biol. Assoc. U.K. 55:925-
931.
Daniels, R. A. 1978. Nesting behaviour of Harpagifer bispinis in
Arthur Harbour, Antarctic Peninsula. J. Fish Biol. 12:465-474.
. 1981. Cryoihenia peninsulae, a new genus and species of no-
totheniid fish from the Antarctic Peninsula. Copeia 1981:558-562.
Danil'chenko, p. G. 1947. Phylogenelic relations between the genera
Palaeogadus and Merluccius. Dokl. Akad. Nauk SSSR 58:659-662.
(In Russian; transl. by J. H. Slep, 1967, U.S. Bur. Commer. Fish.,
Washington, D.C., Transl. 46.)
. 1950. The genus Palaeogadus and its development. Tr. Pa-
leontol. Inst. Akad. Nauk SSSR 25:3-25. (In Russian, transl. by J.
H. Slep, 1968, U.S. Bur. Commer. Fish., Washington. DC, Transl.
53.)
. 1960. Bony fishes of the Maikop Deposits of the Caucasus.
Tr. Paleontol. Inst. Akad. Nauk SSSR 78:1-247.
Dannevig, a. 1919. Fish eggs and larvae, p. 1-80. In: Canadian
fisheries expedition 1914-15 in the Gulf of St. Lawrence and At-
lantic waters of Canada. J. Hjort (ed.). Can. Nav. Serv., King's
Printer, Ottawa.
David, A. 1961. Notes on the embryonic and larval development of
the "GooT\c\\." — Baganus baganus (Hamilton). J. Zool. Soc. India
13:194-205.
David, L. R. 1939. Embryonic and early larval stages of the grunion,
Leuresthes tenuis, and the sculpin, Scorpaena guttata. Copeia 1939:
79-8 1 .
Davis, J. R., and D. E. Louder. 1969. Life history of ecology of
Menidia extensa. Trans. Am. Fish. Soc. 98:466-472.
Davis, W. P. 1966. A review of the dragonets (Pisces: Callionymidae)
of the western Atlantic. Bull. Mar. Sci. 16:834-862.
Davy, B. 1972. A review of the lantemfish genus Taaningichthys
(family Myctophidae) with the description of a new species. Fish.
Bull. U.S. 70:67-78.
Dawson, C. E. 1971a. Occurrence and description of prejuvenile and
early juvenile Gulf of Mexico cobia, Rachycentron canadum. Co-
peia 1971:65-71.
. 1971b. Records of the pearlfish, Carapus bennudensis. in the
Northern Gulf of Mexico and of a new host species. Copeia 1971:
730-731.
. 1973. Indo-Pacific distribution of microdesmid fishes (Go-
bioidea). J. Mar. Biol. Assoc. India 15(l):3l8-322.
. 1 974a. Studies on eastern Pacific sand stargazers (Pisces: Dac-
tyloscopidae). 1 . Platygillelhis new genus, with descriptions of new
species. Copeia 1974:39-55.
. 1974b. A review of the Microdesmidae (Pisces: Gobioidea).
1 . Cerdale and Clarkichthys with descriptions of three new species.
Copeia 1974:409-448.
. 1975. Studies on eastern Pacific sand stargazers (Pisces: Dac-
tyloscopidae). 2. Genus Dactyloscopus. with descriptions of new
species and subspecies. Nat. Hist. Mus. Los Ang. Cty. Sci. Bull. 22.
. 1976. Studies on eastern Pacific sand stargazers. 3. Dactylag-
nus and Myxodagnus. with description of a new species and sub-
species. Copeia 1976:13-43.
. 1982. Atlantic sand stargazers (Pisces: Dactyloscopidae), with
descriptions of one new genus and seven new species. Bull. Mar.
Sci. 32:14-85.
, F. Yasuda and C. Imai. 1979. Elongate dermal appendages
in species of Yozia (Syngnathidae) with remarks on Trachyrham-
phus. Jpn. J. Ichthyol. 25:244-250.
Dean, D. M. 1972. Osteology and systematics of the echeline worm
eels of the Atlantic Ocean (Pisces, Anguilliformes). Unpubl. PhD
dissertation. Univ. Miami, Coral Gables, FL.
Dearborn, J. H. 1965. Reproduction in the nototheniid fish Tre-
matoinus bernacchii Boulenger at McMurdo Sound, Antarctica.
Copeia 1965:302-307.
de AzEVEDO, P., and A. L. Gomes. 1942. Contribuifao ao estudo da
biologia da traira Hoplias malabarica (Bloch, 1 749). Bol. Ind. Anim.
5:15-64.
LITERATURE CITED
699
, M. ViANNA DiAS AND B. BoRGES ViEiRA. 1938. Biologia do
saguiru. Mem. Inst. Oswaldo Cruz Rio de J. 33(4):48 1-554.
DE Beaufort, L. F., and W. M. Chapman. 1951. The fishes of the
IndoAustralian Archipelago. E. J. Brill, Ltd., Leiden.
DeBeer, G. R. 1937. Thedevelopment of the vertebrate skull. Oxford
Univ. Press, London.
. 1951. Embryos and ancestors (revised edition). Oxford at the
Clarendon Press, London.
DE BuEN, F. 1940. Huevos, crias, larvas yjovenes de C/!/ro.sro/«a del
Lago de Patzcuaro. Trab. Estac. Limnol. Patzcuaro 3:1-14.
DE CiECHOMSKi, J. D. 1965. Observaciones sobre la reproduccion,
desarrollo embrionario y larval de la anchoita argentina (£'igraw//.s
anchoila). Bol. Inst. Biol. Mar. 9:1-29.
. 1968. Huevos y larvas de tres especies de peces marinos,
Anchoa marinii, Brevoortia aurea y Prionolus nudigula de la zona
de Mar del Plata. Bol. Inst. Biol. Mar. 17:1-28.
. 1972, Desarrollo embrionario y larval del comalito Aus-
troatherina incisa (Jenyns, 1942) Marrero, 1950. Pisces Atherini-
dae. An. Soc. Cient. Argent. 193:273-281.
. 1981. Ictioplancton. In: Atlas del zooplancton del Atlantico
suboccidental y metodos de trabajo con zooplancton marine. D.
Botovskoy (ed.). Institute Nacional de Investigacion y Desarollo
Pesquero (INIDEP), Ministeno de Comercio e Intereses Maritimos,
Republica Argentina.
, andC. I. BooMAN. 1981. Descripcion deembrionesy de areas
de reproduccion de los granaderos Macrourus whitsoni y Coelo-
rhynchus fascialus. de la polaca Micromesistius aiisl rails y del ba-
calao austral Salilota australis en la zona patagonica y fueguina del
Atlantico sudoccidental. Physis Secc. A Oceanos Org. 40 (98):5-
14.
, AND G. Weiss. 1973. Reproduccion, desarrollo embrionario
y larval del surel, Trachurus picluratus australis (Pisces), en el area
del Mar del Plata y zonas adyacentes. Physis Secc. A Oceanos Org.
32:85-93.
, AND . 1974. Distribucion de huevos y larvas de la mer-
luza, Merlucclus merluccius hubhsi. en las aguas de la plataforma
de la Argentina y Uruguay en relacion con la anchoita, Engraulis
anchoita. y las condiciones ambientales. Physis Secc. A Oceanos
Org. 33(86): 1 85-1 98.
, AND . 1976. Desarrollo y distribucion de postlarvas del
robalo Eleglnops maclovinus (Valenciennes, 1830) Dollo 1904; de
la merluza negra Dissostlchus eleglnoides Smitt, 1899 y de las no-
totenias Nolothenia spp. Pisces. Nototheniidae. Physis Secc. A
Oceanos Org. 35(9 1 ): 1 1 5-1 25.
DeGaetani, D. 1926. Contribute alia conoscenzadello sviluppopost-
larvale di Mora medllerranea Risso. Mem. R. Com. Talassogr. Ital.
129.
. 1928. Uova y larve de Physiculus^ dalwlgkii Kp. Mem. R.
Com. Tallasogr. Ital. 137.
. 1937. Contribute alia cenoscenza delle sviluppo post-em-
brionale in Apogon Imberbis L. Mem. R. Com. Talassogr. Ital. 243.
— . 1940. Stadi larvali e giovanili di L(c/»a g/a!(ca Risso. Mem.
R. Com. Talassogr. Ital. 270:1-15.
Degens, E. T, W. G. Deuser and R. L. Haedrich. 1969. Molecular
structure and composition of fish otoliths. Mar. Biol. (Berl.) 2:105-
113.
DE Jager, B. v. D. 1955. The South Afncan pilchard (Sardinops ocel-
toa)— The development of the snoek ( Thyrsnes alun). a fish pred-
ator of the pilchard. Invest. Rep. Div. Fish. S. Afr. 19:1-16.
Dekhnik, T. V. 1959. Material on spawning and development of
certain far eastern flatfishes. Akad. Nauk. SSSR, Zool. Inst. Issled.
Dafnevost. Morei SSSR 6:109-131. (In Russian.)
. 1973. Iktioplankton chemogo moria. (Ichthyoplankton of the
Black Sea). Naukova Dumka, Kiev.
, M. HiiARES AND D. Salabariya. 1966. Distribution of pelagic
eggs and larvae of fishes in the coastal waters of Cuba. Inst. Biol.
Southern Seas, Acad. Sci. USSR 1 ( 1 ): 1 38-1 6 1 . (English translation
1973, U.S. Dep. Commerce, TT70-57762 Investigations of the
Central American Seas).
, AND V. I. SiNYUKovA. 1966. Distribution of pelagic fish eggs
and larvae in the Mediterranean Sea. Part II. On the reproduction
and ecology of larvae in Mediterranean Myclophidae. In: Studies
on the plankton of southern seas. Nauka, Moskow. (In Russian;
English transl. No. 55. Bur. Commer. Fish., Syst. Lab., Washing-
ton.)
DeLacy, a. C, and B. S. Batts. 1963. Possible heterogeneity in the
Columbia River smelt. Univ. Wash. Fish. Res. Inst. Circ, No. 198.
, C. R. HiTZ, and R. L. Dryfoos. 1964. Maturation, gestation
and birth of rockfish (Sebastodes) from Washington and adjacent
waters. Wash. Dep. Fish. Fish. Res. Pap. 2(3):51-67.
DeLamater, E, D.. and W. R. Courtenay, Jr. 1973a. Studies on
the scale structure of flat-fishes I. the genus Trinectes. with notes
on related forms. Proc. Southeast Game Fish Comm. 1973:591-
608.
, AND . 1973b. Variations in the structure of the lateral
line canal on scales of teleostean fishes. Z. Morphol. Tiere 75:259-
326.
, AND . 1974. Fish scales as seen by scanning electron
microscopy. Fla. Sci. 37:141-149.
Delsman, H. C. 1921. Fish eggs and larvae from the Java Sea. 1.
Fistularia serrata Cuv. Treubia 2:92-108.
. 1923. Fish eggs and larvae from the Java Sea. 2. C/;;rofe«/rw.s
dorab. Treubia 3:38-46.
. 1925. Fish eggs and larvae from the Java Sea. 4. Dussumieria
haseltii. Treubia 6:297-307.
. 1926a. Fish eggs and larvae from the Java Sea. 5. Caran.x
kurra. macrosorna. and crumenophthalmus. Treubia 8: 1 99-2 1 1 .
. 1 926b. Fish eggs and larvae from the Java Sea. 7. The genus
Clupea. Treubia 8:218-239.
. 1926c. Fish eggs and larvae from the Java Sea. 8. Dorosoma
chacunda (H.B.). Treubia 8:389-394.
. 1926d. Fish eggs and larvae from the Java Sea. 10. On a few
larvae of empang fishes. Treubia 8:400-412.
. 1927. Fish eggs and larvae from the Java Sea. 1 1. The genus
Trichiurus. Treubia 9:338-351.
— . 1929a. Fish eggs and larvae from the Java Sea. 12. The genus
Engraulis. Treubia 11:275-281.
— . 1929b. Fish eggs and larvae from the Java Sea. 13. Chanos
chanos (Forsk). Treubia 1 1:281-286.
— . 1929c. The study of pelagic fish eggs. Proc. Pac. Sci. Congr..
4th Ses. 3:61-67.
. 1930a. Fish eggs and larvae from the Java Sea. 14. The genus
Pellona. Treubia 12:37-46.
. 1930b. Fish eggs and larvae from the Java Sea. 15. On Chl-
rocentrus hypselosoma and dorab. Treubia 1 2:46-50.
. 1930c. Fisheggsandlarvaefrom the JavaSea. 16. .-t/jpp/^pr/o/!
perculac.\. Treubia 12:367-370.
— . 1930d. The study of pelagic fish-eggs. Proc. Fourth Pac. Sci.
Congr. (1929) 3:61-67.
— . 1931. Fish eggs and larvae from the Java Sea. 1 7. The genus
Stolephorus. Treubia 13:217-243.
— . 1 932a. Fish eggs and larvae from the Java Sea. 1 9. The genus
Selipmna. Treubia 14:109-114.
— . 1932b. Fish eggs and larvae from the Java Sea. 20. The genus
Coilia. Treubia 14:1 14-1 16.
— . 1933a. Fish eggs and larvae from the Java Sea. 22. C/!*^^^);^^^
lile(C.V.). Treubia 14:247-249.
— . 1933b. Fish eggs and larvae of the Java Sea. 23. A few more
Clupea-eggs. Treubia 14:249-254.
— . 1938. Fish eggs and larvae from the Java Sea. 24. Mycto-
phoidea. Treubia 16:415-420.
, AND J. D. G. Hardenberg. 1934. De Indische Zeevisschen
en Zeevischeriji, p. 330-333. N. V. Boekhandel en Drukkerij Visser
Co., Batavia.
DeMartini, E. E. 1976. The adaptive significance of territoriality and
egg cannibalism in the painted greenling, Oxylebius pictus Gill, a
northeastern Pacific marine fish. Unpubl. PhD thesis. Univ. Wash-
ington, Seattle.
700
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
. 1978. Apparent paternal care in Lipans fucensis (Pisces: Cy-
clopteridae). Copeia 1978:537-539.
Demir, M. 1961. On the eggs and larvae of rrac/iwnM ?rac/n<n« (L.)
and Trachurus mediterranus (Stohrr.) from the Sea of Marmara
and the Black Sea. Rapp. P. V. Reun. Comm. Int. Explor. Sci. Mer
Mediterr. Monaco 16:317-320.
Demir,N. 1972. The abundance and distribution ofthe eggs and larvae
of some teleost fishes off Plymouth in 1969 and 1970. II. The
postlarvae of Callionymus. J. Mar. Biol. Assoc. U.K. 52:997-1010.
. 1976. Callionymidaeof the northeastern North Atlantic. Fich.
Ident. Zooplanclon 148:1-5.
. 1982. On postlarvae and pelagic juveniles of rocklings, Ga;(y-
ropsarus mediterranem and G. hiscayensis. J. Mar. Biol. Assoc.
U.K. 62:647-665.
Dessier, a. 1969. Note sur les stades larvaires et post-larvaires d'
Ilisha africana (Bloch, 1795) (Pisces, Clupeidae). Cah. ORSTOM
Ser. Oceanogr. 8(4):21-27.
DE Sylva, D. P. 1963. Systematics and life history of the great bar-
racuda, Sphyraena barracuda (Walbaum). Stud. Trop. Oceanogr.
(Miami).
. 1975. Barracudas (Pisces: Sphyraenidae) of the Indian Ocean
and adjacent seas — a preliminary review of their systematics and
ecology. J. Mar. Biol. Assoc. India 15(l):74-94.
, AND W. N. EscHMEYER. 1977. Systematics and biology of the
deep-sea fish family Gibberichthyidae, a senior synonym of the
family Kasidoroidae. Proc. Calif Acad. Sci., Ser. 4, 41:215-231.
AND L. N. ScoTTON. 1972. Larvae of deep-sea fishes (Sto-
miatoidea) from Biscayne Bay, Florida, USA, and their ecological
significance. Mar. Biol. (Bed.) 12:122-128.
Detwyler, R. 1963. Some aspects of the biology of the seasnail,
Liparis atlanticus (Jordan and Evermann). Unpubl. PhD disser-
tation. Univ. of New Hampshire.
Deuel, D. G., J. R. Clark and A. J. Mansueti. 1966. Description
of embryonic and early larval stages of the bluefish, Pomatomus
saltatnx. Trans. Am. Fish. Soc. 95:264-271.
Devanesan, D. W. 1937. On the spawning habits and the hatching
experiments with the eggs of Hemirhamphus georgii Cuv. et Val.
C. R. 12th Congr. Intematl. Zool. (Lisbon) 3:2175-2181.
Devaraj, K. v., T. J. Varchese and G. P. S. Rao. 1972. Induced
breeding ofthe freshwater catfish, Clarias balrachus (,Lmn.)by using
pituitary glands from manne catfish. Curr. Sci. 41:868-870.
Devi, C. B. L. 1969. Occurrence o{ larvae of Psctidorlionihiis elevatus
Ogilby (Heterosomata-Pisces) along the south-west coast of India.
Proc. Indian Acad. Sci. Sect. B: 178-1 86.
DeWitt, H. H. 1962a. A new genus and species of zoarcid fish from
McMurdo Sound. Antarctica. Copeia 1962:819-826.
. 1962b. A new Antarctic nototheniid fish with notes on two
recently described nototheniiformes. Copeia 1962:826-833.
. 1964a. A revision of the Antarctic genus /^acYnv/rM (Pisces,
Bathydracomdae). Copeia 1964:496-506.
. 1 964b. A redescription of Pathogenis anlarctica. with remarks
on the genus Treinatomus (Pisces. Nototheniidae). Copeia 1964:
683-686.
. 1970. A revision of the fishes of the genus Notothenia from
the New Zealand region, including Macquarie Island. Proc. Calif
Acad. Sci. 38(16):299-340.
. 1971. Coastal and deep-water benthic fishes of the Antarctic.
Am. Geogr. Soc, Antarct. Map Folio Ser., Folio 15.
. 1977. A new genus and species of eelpout (Pisces, Zoarcidae)
from the Gulf of Mexico. Fish. Bull. U.S. 75:789-793.
, AND J. C. Hureau. 1979. Fishes collected during "Hero"
Cruise 72-2 in the Palmer Archipelago Antarctica with the descrip-
tions of two new genera and three new species. Bull. Mus. Natl.
Hist. Nat. 4e Ser. I, 1979, Sec. A (3):775-820.
, AND J. C. Tyler. I960. Fishes from the Stanford Antarctic
da lumbil (Bloch) off southwest coast of India, p. 460-473. In:
Proc. Symp. Warm Water Zooplankton. Natl. Inst. Oceanogr.
DiMkhele, L., AND M. H. Taylor. 1980. The environmental control
of hatching in Fundidus heteroclitus. J. Exp. Zool. 214:181-187.
DiNGERKUS, G., AND L. D. Uhler. 1977. Enzyme clearing of alcian
blue stained whole small vertebrates for demonstration of cartilage.
Stain Technol. 52:229-232.
DiTTY, J. G., AND F. M. Truesdale. 1983. Comparative larval de-
velopment of Peprilus burti. P. triacanlhus and P. paru (Pisces:
Stromateidae) from the western North Atlantic. Copeia 1983:397-
406.
DoBBS, G. H. 1974. Soft tissues of marine teleosts, p. 72-82. In:
Principles and techniques of scanning electron microscopy. M. A.
Hayat (ed.). Van Nostrand Reinhold Co., New York.
. 1975. Scanning electron microscopy of inlraovarian embryos
ofthe viviparous teleost. Micrometrus minimus (Gibbons), (Per-
ciformes: Embiotocidae). J. Fish Biol. 7:209-214.
Doi, M., AND O. Okamura. 1983. The chlorophthalmid fishes col-
lected from the western Pacific. Abstract of paper presented at the
15th Ann. Meet, ofthe Ichthyol. Soc. Jpn., 1983.
DooLEY, J. S. 1978. Systematics and biology ofthe tilcfishes (Perci-
formes: Branchiostegidae and Malacanthidae) with descriptions of
two new species. NOAA Tech. Rep. NMFS Circ. 411.
DoROFE^EVA, Y. A., Y. A. Zinov'yev, v. a. Klyukanov, Y. S.
Reshetnikov, K. a. Savvaitova and G. K. Shaposhnikova.
1 980. The present state of research into the phylogeny and classi-
fication of Salmonoidei. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
20(5): 1-20.
DoROSHEv S. 1 . 1 970. Biological features ofthe eggs, larvae, and young
ofthe striped bass (Roccus saxatilis (Walbaum)) in connection with
the problem of its acclimitization in the U.S.S.R. J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 10:235-248.
, AND J. W. Cornacc HiA. 1979. Initial swim bladder inflation
in the larvae of TUapia mossambica (Peters) and Morone saxatdis
(Walbaum). Aquaculture 16:57-66.
, AND K. HotiAN. 1981. Initial swim bladder inflation
in the larvae of physoclistous fishes and its importance for larval
culture. Rapp. P V. Reun. Cons. Int. Explor. Mer 178:495-500.
DoTSU, Y. 1954. On the life history of a goby Chaenogobius castanea
O'Shaughnessy. Contnb. Biol. Lab. Sci. Soc. China 1 1(6):133-I38.
. 1957. On the bionomics and life history of the eel-like goby,
Odontamblyopusrubicundu5(YiZLVn\\lon). Sci. Bull. Fac. Agric. Kyu-
shu Univ. 16(1):101-1I0.
. 1958. The bionomics and larvae ofthe goboid fishes, C/p/io(n'-
pauchen microcepholus (Bleeker) and Taenioides cirratus (Blyth).
Sci. Bull. Fac. Agric. Kyushu Univ. 16(3):37 1-380.
. 1979. Eggs, larvae, and juveniles of gobioid fishes. Kaiyo
Kaguku (Mar. Sci.) 11:111-1 16. (In Japanese.)
. 1982. The early life history of the combtooth blenny, £)aMo/i
Irossulus and spawnings of the laboratory-reared fish about three
months after hatching. Bull. Fac. Fish. Nagasaki Univ. 52:19-27.
(In Japanese, English summary.)
, AND S. FujiTA. 1959. The bionomics and life history ofthe
eleotrid fish. Eleolris oxycephata Temminck and Schlegel. Bull. Fac.
Fish. Nagasaki Univ. 8:191-195.
, andS. MiTO. 1955. On the breeding habits, larvae, and young
of a goby, Acanthogohius flavimanus (Temminck et Schlegel). Jpn.
J. Ichthyol. 4:152-161.
, AND . 1963. The bionomics and life history of the
eleotrid fish, Asterropteryx semipunclalus Ruppell. Bull. Fac. Fish.
Nagasaki Univ. 15:1-3.
, AND S. MoRRicHi. 1980. The life history of the blcnniid fish.
Biological Research Program 1958-1959. Stanford Ichthyol. Bull.
7(4):162-I99.
DiLEEP,M.P. 1977. The larval development and distribution of S'lJi/n-
BtenniusyalabeiOorAanelSnyAer). Bull. Fac. Fish. Nagasaki Univ.
49:17-24. (In Japanese, English summary.)
— , AND T. OoTA. 1973. The life history of the blenniid fish,
Omobranchus loxozonus. Bull. Fac. Fish. Nagasaki Univ. 36:13-
22. (In Japanese, English summary.)
— , AND M. SHiociAK.1. 1971. Larvae, juvemlcs and young of the
gobiid fish, Astrabe lactisella. Jpn. J. Ichthyol. 18:182-186.
LITERATURE CITED
701
DovEL, W. L. 1963. Larval development of clingfish. GoWmo.v i^rw-
mosiis. 4.0 to 12.0 millimeters in total length. Chesapeake Sci.
4:161-166.
Dryfoos, R. L. 1965. The life histor>' and ecology of the longfin smelt
in Lake Washington. Unpubl. PhD dissertation. University of
Washington.
DiiDNiK, Y. I. 1977. A contnbution to the biology of larvae and fry
of morwongs (Pisces, Cheilodactylidae) of the Atlantic Ocean. Zool.
Zh. 56(11);1658-1667.
DuMONT, J. N., AND A. R. Brummett. 1980. The vitelline envelope,
chorion, and micropyle of Fundulus heteroclitus egg,s. Gamete Res.
3:25-44.
Dunbar, M. J. 1947. Marine young fish from the Canadian eastern
Arctic. Bull. Fish. Res. Board Can. 73:1-1 1.
DiiNCKER, G., andE. MoHR. 1939. Revisionder Ammodytidae. Mitt.
Zool. Mus. Berl. 24(I):3-3I.
DuNKELBiiRGER, D. G., J. M. Dean AND N. Watabe. 1980. The ul-
trastructuie of the otolithic membrane and otolith in the juvenile
mummichog, Fundulus heterocliliis. ]. Morphol. 163:367-377.
Dunn. J. R. 1983a. Development and distribution of the young of
northern smoothtongue, Leuroglossus schmidli (Bathylagidae). in
the northeast Pacific, with comments on the systemalics of the
Genus Leuroglossus G\\hen. Fish. Bull. U.S. 81:23-40.
. 1983b. The utility of developmental osteology in ta,\onomic
and systematic studies of teleost larvae: a review. NOAA Tech.
Rep. NMFS Circ 450.
, AND B. M. ViNTER. 1984. Development of larvae of saffron
cod, Eleginus gracilis, with comments on the identification of gadid
larvae in Pacific and Arctic waters contiguous to Canada and Alas-
ka. Can. J. Fish. Aquat. Sci. 41:304-318.
, AND . MS. The developmental osteology of the North
Pacific gadid fish, Theragra chalcogramma. Unpubl. manuscr..
Northwest and Alaska Fisheries Center, National Marine Fisheries
Service, NOAA, 2725 Montlake Blvd. E., Seattle, WA 981 12.
DuRAND.J, R., andG. LouBENS. 1971. Developpementembryonnaire
et larvaire A'Alestes baremoze. Cah. ORSTOM. Ser. Hydrobiol.
5:137-145.
DuTT, S., AND V. V. Rao. 1960. On the breeding habits and eariy
developmental stages of Peirosciries bipunclala (Day). ,1. Zool. Soc.
India 12:158-161.
D'ViNCENT, S.. H. G. MosER AND E. H. Ahl-STROM. 1980. Descnption
of the larvae and early juveniles of the Pacific butlerfish, Peprilus
simillimus (Family Stromateidae). Calif Coop. Oceanic Fish. In-
vest. Rep. 21:172-179.
Dye, J. E. 1980. The production and eflicient use of freshly hatched
brine shrimp nauplii (Anemia) in the larval rearing of marine fish
at the hatcheries of the Bntish Whitcfish Authority. In: The Bnne
Shrimp Anemia. Vol. 3. G. Persoone, P. ^orgeloos, O. Roels and
E. Jaspers (eds.). Universa Press, Wetleren, Belgium.
Eat(in, T. H. 1953. Pygmy sunfishes and the meaning of pygmyism.
J. Elisha Mitchell Sci. Soc. 69:98-99.
. 1956. Notes on the olfactory organs in Centrarchidae. Copeia
1956:196-199.
Ebeleng, A. W. 1957. The dentition of eastern Pacific mullets, with
special reference to adaptation and taxonomy. Copeia 1957:173-
185.
. 1961. .Wi(?;7 ga/apagCT!i;.s, a new mullet from the Galapagos
Islands, with notes on related species and a key to the Mugilidae
of the eastern Pacific. Copeia 1961:295-304.
. 1962. Melamphaidac. I. Systematics and zoogeography of the
species in the bathypelagic fish genus .Melamphaes Gunther. Dana-
Rep. Carlsberg Found. 58.
, AND W. Weed. 1963. Melamphaidac. 111. Systematics and
distribution of the species in the bathypelagic fish genus Scopelo-
gadus Vaillant. Dana-Rep. Carlsberg Found. 60.
, AND . 1973. Order Xenoberyces (Sephanoberyci-
(Paralepididae, Pisces). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
18:500-504.
. 1983a. Atlas offish larvae from the Southern Ocean. Cybium,
3e Ser. 7(2): 1-74.
. 1983b. Description ofeggs and larvae of A/«rae«o/ep/im/cTOpj
Lonnberg (Muraenolepidae) from the Scotia Sea. J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 23(1):I39-I42.
, A. N. KoTLVAR AND L. A. LisovENKO. 1981. Sex cycles, fe-
cundity, and distribution of "limbless" soles of family Bothidae
from the southwestern Atlantic. .1. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.)2l(6):89-99.
, AND L. A. LrsovENKO. 1970. Morphological features of in-
traovarian and pelagic larvae of some Sebastodes species inhabiting
the Gulf of Alaska, p. 267-286. In: Soviet fisheries investigations
in the northeast Pacific. Part V. P. A. Moiseev (ed.). (Transl. from
Russian; available from: NTIS. Springfield, VA TT7 1-50 127.)
Ege, V. 1918. Stomiatidae. Rep. Dan. Oceanogr. Exped. Mediterr.
2(A.4).
. 1930. Sudidae(Para/e/)K). Rep. Dan. Oceanogr. Exped. Med-
iterr. 2(A.13).
. 1953. Paralepididae I (Para/epii and Le5/;V/;H/H). Taxonomy,
ontogeny, phylogeny and distribution. Dana-Rep. Carlsberg Found.
40.
. 1957. Paralepididae II (Macroparalepis). Taxonomy, ontog-
eny, phylogeny and distribution. Dana-Rep. Carlsberg Found. 43.
1958. Omosudis Gunther, bathypelagic genus offish. Dana-
formes), p. 397-478. In: Fishes of the western North Atlantic. D.
M. Cohen (ed.). Mem. Sears Found. Mar. Res. No. 1 (Pt. 6).
Efremenko, V. N. 1978. A description of the larvae of A'ow/cp/sroari/
Rep. Carlsberg Found. 47.
Eggert, B, 1932. Zur Kenntnis der Biologic, der sekundaren Ge-
schlechtsmerkmale und des Eies von Blennius pavo (Risso). Z. Mor-
phol. Oekol. Tiere 24:682-703.
Ehrenbaum, E. 1 894. Beitrage zur Naturgeschichte einiger Elbflsche.
(Osmerus epertanus L., Clupea finia Cuv., .-icerina cernua L., Aci-
penserslunoL.). Wiss. Meersunters. (Helgol.) l(Pt.I ):35-82. (Cited
m: W. C. Kendall. 1 927. The smelts. Bull. Bur. Fish. Wash. 42( 1 926):
217-375.)
. 1897. Eier und Larven von Fischen der Deutschen Buchl.
Wiss. Meeresunters. (Helgol.) 2:253-324.
. 1904. Eier und Larven von Fischen der Deutschen Bucht. III.
Fische mit festsitzenden Eiem. Wiss. Meeresunters (Helgol.) N.F.
6:127-200.
. 1905-1909. Eier und Larven von Fischen. Nord. Plankton 1
and 2.
, AND S. Strodtmann. 1904. Eier und Jugendformen der Ost-
seefische. Wiss. Meeresunters. (Helgol.) 6:57-126.
Ehrich, S. 1976. Zur Taxonomie, Okologie und Wachstum von Ma-
crorhamphosus scnlopa.x (Linnaeus, 1758) (Pisces, Syngnathi-
formes) aus dem subtropischen Nordostatlantik, Ber. Dtsch. Wiss.
Komm. Meeresforsch. 24:523-537.
Ehrlich, K. F., J. H. S. Blaxter and R. Pemberton. 1976. Mor-
phological and histological changes during the growth and starva-
tion of herring and plaice larvae. Mar. Biol. 35:105-118.
Eigenmann, C. H. 1902. The egg and development of the conger eel.
Bull. U.S. Fish Comm. 21:37-44.
. 1928. The freshwater fishes of Chile. Mem. Natl. Acad. Sci.
22:1-80.
EINARSSON, H. 1951. The post-larval stages of sandeels in Faroe, Ice-
land, and W-Greenland waters. Acta Nat. Isl. l(7):5-75.
. 1955. On the post-larval stages of .■lm/)Joa'i7M/a/!fea(Cuvier).
Acta Nat. Isl. 2(l):3-7.
, and B. Rojas de Mendioea. 1963. Descripcion de huevos y
larvas de anchoveta Peruana {Engraiilis ringens }.). Bol. Inst. Mar
Peru(Callao) 1.
Eldred, B. 1967a. Larval bonefish, .Ithula vulpcs (Linnaeus. 1758)
(Albulidae) in Flonda and adjacent waters. Fla. Board Conserv.
Mar. Res. Lab. Lead. Ser. 4 (Pt. 1, No. 3).
. 1967b. Larval tarpon, Megalops allanlicus Valenciennes
(Megalopidae) in Florida waters. Fla. Board Mar. Res. Lab. Leafi.
Ser. 4 (Pt. 1, No. 4).
. 1968. The larval development and taxonomy of the pygmy
702
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
moray eel, Anarchias yoshiaeKanazcLVja 1952. Fla. Board Conserv.
Mar. Res. Lab. Leafl. Ser. 4 (Pt. 1. No. 10).
. 1969. Embryology and development of the Blackedge moray
Gymnolhorax nigromarginatus (Girard. 1859). Fla. Board Conserv.
Mar. Res. Lab. Leafl. Ser. 4 (Pt. 1, No. 13).
. 1972. Note on larval tarpon, A/e.fa/opi a//an;;cMi (Megalopi-
dae), in the Florida Straits. Fla. Dep. Nat. Resour. Mar. Res. Lab.
Leafl. Ser. 4 (Pt. 1. No. 22).
, AND W. G. Lyons. 1966. Larval ladyfish, Elops saurus Lm-
naeus 1766, (Elopidae) in Florida and adjacent waters. Fla. Board
Conserv. Mar. Res. Lab., Leafl. Ser. 4 (Pt. 1. No. 2).
Eldridge, M. B. 1970. Larval fish survey of Humboldt Bay. Unpubl.
MS thesis. Humboldt State Univ., Areata, CA.
. 1975. Early larvae of the diamond turbot. Hypsopsetia giil-
tulata. Calif Fish Game 61:26-34.
Elston, R., L. Corazza and J. G. Nickhm. 1981. Morphology and
development of the olfactory organ in larval walleye, Stizostedion
vilreum. Copeia 1981:890-893.
Emery, A. 1973. Biology survey — summer e.xpedition. //i. Arctic Div-
ing. Adv. Comm. Nor. Dev., No. 60°, Govt. Canada, Ottawa.
Emery, C. 1879. Le metamorfosi del Trachipterus taenia. Mitt. Zool.
Sta. Neapel 1:58-588.
. 1880. Le specie del genere Fierasfer nel golfo di Napoli e
regione limitrofe. Fauna Flora Golf Neapel 2:1-76.
. 1886. Contribuzioni all'ittiologia. XVIL Larva di genere ig-
noto. Mitt. Zool. Sta. Neapel 6:149-164.
Eng, C. T. 1969. Early development of the glassfish [Amhassis kopsii
Elk). J. Singapore Natl. Acad. Sci. l(2):37-44.
Engen, P. C. 1968. Organogenesis in the walleye surfperch, Hyper-
prosopon argenleum (Gibbons). Calif Fish Game 54:156-169.
Erdahl, D. A., AND E. F. Grahm. 1980. Preservation of gametes of
freshwater fish, p. 317-336. In: Proc. 9th International Congress
on Animal Reproduction and A.L Vol. II, Round Tables, Invited
Guests. Madrid, Spain.
EscHMEVER, W. N.. and B. B. Collette. 1966. The scorpionfish
subfamily Setarchinae, including the genus Ectreposebastes. Bull.
Mar. Sci. 16:349-375.
, E. Herald and H. Hammann. 1983. A field guide to Pacific
Coast fishes of North America. Peterson Field Guide Series 28.
Houghton Mifllin Co., Boston.
-, and J. C. HuREAU. 1971. Sehasles mouchezi. a senior syn-
onym of Helicoleniis tnstanensis. with comments on Sebastes ca-
pensis and zoogeographical considerations. Copeia 1971:576-579.
EsPiNOSA, F. A. 1968. Spawning periodicity and fecundity of C/'e«;f/(-
thys baileyi, a fish endemic to Nevada. Unpubl. MS thesis. Nevada
South. Univ., Las Vegas.
EsTABROOK. G. F., C. S. Johnson, Jr. and F. R. McMoRRis. 1976. A
mathematical foundation for the analysisofcladisticcharacter com-
patibility. Math. Biosci. 29:181-187.
Evans, J. W. 1962. Normal stages of the early development of the
flying fish, Hirundichlhys ajfinis (Gunther). Bull. Mar. Sci. Gulf
Carib. ll(4):483-502.
Evans, R. R. 1973. The swimbladder and associated structures in
western .Atlantic sea robins (Triglidae). Copeia 1973:315-321.
EvERMANN, B. W., AND W. C. KENDALL. 1898. Descriptions of new
or little known genera and species of fishes from the United States.
Bull. U.S. Fish. Comm. 17:125-133.
EvERSON, I. 1968. Larval stages of certain Antarctic fishes. Bull. Br.
Antarct. Surv. 16:65-70.
. 1970. Reproduction in No/oJ/ien/a neg/ec(a Nybelin. Bull. Br.
Antarct. Surv. 33:81-92.
EvsEENKO. S. A. 1977. Larval Engyophrys senlus Ginsburg. 1933
(Pisces, Bolhidae) from the American Mediterranean Sea. Coop.
Invest. Carib. Adjac. Reg.-II. FAO Fish. Rep. 200:171-185.
. 1979. Larvae ofthe flounder CR7op5c»a Gill. 1888 (Bothidae,
Pisces) from the north-western Atlantic. Biol. Morva (Vladivost.)
2:67-75.
. 1981. On the sinistral flatfish larvae (Scophthalmidae, Both-
idae, Pisces) from the west Atlantic, p. 593-594. In: The early life
history of fish. Recent studies. R. Lasker and K. Sherman (eds.).
Rapp. P. V. Reun. Cons. Int. Explor. Mer 178.
. 1982a. Eco-morphological peculanties of the early life history
of flatfishes of the western North Atlantic. Proc. Ichthyoplankton
and Nekton of the World Ocean. P. P. Shirshov Inst. Oceanol.
Acad. Sci. USSR. 1 18:43-84.
. 1982b. Ichthyoplankton of slope and Gulf Stream waters off
Nova Scotia in late autumn 1974. J. Northwest Atl. Fish. Sci.
3:127-139.
Faber, D. J. 1970. Ecology of larval whitefishes in Lake Huron, p.
481-500. In: Biology of coregonid fishes. C. C. Lindsey and D. S.
Woods (eds.). Univ. Manitoba Press, Winnipeg.
. 1976. Identification of four northern blennioid fish larvae in
the Canadian Atlantic Ocean (Stichaeidae, Lumpenidae). J. Fish.
Res. Board Can. 33:1798-1802.
. 1982. Fish larvae caught by a light-trap at littoral sites in Lac
Heney, Quebec. 1979 and 1980, p. 42-46. //;. Proc. 5th Annual
Larval Fish Conference. La. Coop. Fish. Res. Unit, Baton Rouge,
LA.
Fabre-Domergue, p., and E. BmxRLX. 1905. Developpement de la
Sole (Solea vulgaris). Trav. Lab. Zool. Mar. Concameau.
Facciola, L. 1882. La forma giovanile del Phycts blennioides. Nota.
Nat. Siciliano (Palermo) 2:25-29.
Fage, L. 1910. Recherches sur les Stades pelagiques de quelques Te-
leosteens de la mer de Nice (Parages de Monaco) et du Golfe du
Lion. Annls. Inst. Oceanogr. (Monaco) 1, Fasc. 7.
. 1918. Shore Fishes. Rep. Dan. Oceanogr. Exped. Mediterr.
2(A.3).
Fahay, M. p. 1975. An annotated list of larval and juvenile fishes
captured with surface-towed meter net in the south Atlantic Bight
during four R/V Dolphin cruises between May 1967 and February
1968. NOAA Tech. Rept. NMFS SSRF 685.
. 1983. Guide to the early stages of marine fishes occurring in
the western North Atlantic Ocean, Cape Hatteras to the southern
Scotian Shelf J. Northwest. .\X\. Fish. Sci. 4.
, AND P. Berrien. 1981. Preliminary description of larval tile-
fish (Lophotatilus chamaeleonticeps). Rapp. P. V. Cons. Int. Explor.
Mer 178:600-602.
, AND C. L. Obenchain. 1978. Leptocephali of the ophichthid
genera .4hlia. Myrophis. Ophichthus. Pisodonophis. Caltechctys.
Letharchus. and .Iptenchtus on the Atlantic continental shelf of the
United States. Bull. Mar. Sci. 28:442-486.
Fahv, W. E. 1982. The influence of temperature change on number
of pectoral fin rays developing in Fundubis majalis (Walbaum). J.
Cons. Int. Explor. Mer 40:21-26.
Fanara, D. M. 1964. Notes on the biology of a salt marsh minnow.
Proc. N.J. Mosq. Exterm. Assoc. 51:152-159.
Fange, R. 1958. The structure and function of the gas-bladder in
Argentina sihts. Q. J. Microsc. Sci. 99:95-102.
F.A.O. 1983. Catches and landings 1981. Yearbook of fishery statis-
tics. 52. FAO, Rome.
Farris, J. S. 1970. Methods for computing Wagner trees. Syst. Zool.
19:83-92.
, A. G. Kllige AND M. J. EchLARDT. 1970. A numerical approach
to phylogenetic systematics. Syst. Zool. 19:172-191.
Farrugio, H. 1977. Clescommentees pour la determination desadultes
et des alevins de Mugilidae de Tunisie. Cybium. 3rd Ser. (2):57-
73.
Fast. T. N. 1 960. Some aspects of the natural history ofStenobraciuus
leucopsariis Eigenmann and Eigenmann. Unpubl. PhD dissertation.
Stanford Univ., Stanford, CA.
Fedorov, V. V. 1975. Description of a new genus and species of
zoarcid fish. Puzanovia rubra, gen. et sp. n. (Pisces, Zoarcidae) from
the northern Pacific Ocean. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
15:527-531.
. 1982. A new ee\pout. Hadropogonichthys tindbergi Fedorov.
gen. et sp. nov. (Zoarcidae), from the bathyal depths of the fourth
LITERATURE CITED
703
Kurile Strait. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 22(5): 16-
23.
-, andA. V. Neyelov. 1978. A new genus and species of eelpoul.
Andriashevia apiera (Perciformes, Zoarcidae). J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 18:846-850.
Fedotov, V. F. 1976. Gadidae of the Palaeogene-Neogene from the
USSR. Tr. Paleonlol. Inst. Akad. Nauk SSSR 157.
Fields, H. M. 1962. Pompanos (rrac/!;/io/;« spp.) of South Atlantic
coast of the United States. U.S. Fish Wildl. Serv. Fish. Bull.
FiERSTiNE, H. L. 1972. The paleontology of billfish — the state of the
art, p. 34-44. In: Proc. Intematl. Billfish Symp., Kailua-Kona,
Hawaii, 9- 1 2 August 1972, Part 2. Review and contributed papers.
R. S. Shomura and F. Williams (eds.). NOAA Tech. Rep. NMFS
SSRF-675.
Fink, S. V., and W. L. Fink. 1981. Interrelationships of the ostar-
iophysan fishes (Teleostei). Zool. J. Linn. Soc. 72:297-353.
Fink, W. L. 1981. Ontogeny and phylogenyoftooth attachment modes
in actinopterygian fishes. J. Morphol. 167:167-184.
. 1982. The conceptual relationship between ontogeny and phy-
logeny. Paleobiology 8:254-264.
, K. E. Hartel, W. G. Saul. E. M. Moon and E. O. Wile'i .
MS. A report on current supplies and practices used in curation
of ichthyoplankton collections. Am. Soc. Ichthyol. & Herpetol.
Mimeo.
, AND S. H. Weitzman. 1982. Relationships of the stomiiform
fishes (Teleostei). with a description of Diplophos. Bull. Mus. Comp.
Zool. Harv. Univ. l50(2):31-93.
FistHER, W. K. 1958a. Primeras fases del desarrollo del blanquillo
{Prolatilus jugidans) Cuv. et Val. (Pisces). Rev. Biol. Mar. 8:3-24.
. 1958b. Huevos, crias y primeras prelarvas de la "anchoveta"
(Engraulis ringens) Jenyns. Rev. Biol. Mar. (Chile) 8:1 1 1-124.
. 1959. Huevos, crias y prelarvas de la merluza. Rev. Biol. Mar.
9(I,2,3):229-249.
. 1963. Die Fische des Brackwassergebietes Lenga bei Concep-
cion (Chile). Int. Rev. Gesamten Hydrobiol. 48:419-51 1.
Fish, M. P. 1928. Contributions to the embryology of the American
eel (Anguilta rostrala Lesueur). Zoologica (N.Y.) 8:289-324.
. 1932. Contributions to the early life histories of sixty-two
species of fishes from Lake Erie and its tributary waters. Bull. Bur.
Fish. Wash. 47(10):293-398.
FisHELSON, L. 1963. Observations on littoral fishes of Israel. II. Larval
development and metamorphosis of Blennius pavo (Risso) (Te-
leostei, Blenniidae). Isr. J. Zool. 12:81-91.
. 1975. Ethology and reproduction of pteroid fishes found in
the Gulf of Aqaba (Red Sea), especially Dendrochirus brachyptenis
(Cuvier), (Pteroidae, Teleostei). Publ. Staz. Zool. Napoli 39 (Suppl.);
635-656.
. 1976. Spawning and larval developrfient of the blenniid fish,
Meiacanthus nigrolineatus from the Red Sea. Copeia 1976:798-
800.
FiTCH, J. E. 1964. The ribbonfishes (Family Trachipteridae) of the
eastern Pacific Ocean, with a description of a new species. Calif
Fish Game 50:228-240.
. 1966. The unicomfish, Eumecichlhys ftski (Gunther) in the
Eastern Tropical Pacific. Calif Fish Game 52:208-210.
. 1967. The tapertail nbbon-fish ( Trachipterus fukuzakii Fitch)
added to the marine fauna of California. Calif Fish Game 53:298-
299.
. 1982. Revision of the eastern Nonh Pacific anthiine basses
(Pisces: Serranidae). Contrib. Sci. (Los Ang.) 339.
, AND L. W. Barker. 1972. The fish family Moridae in the
eastern North Pacific with notes on morid otoliths, caudal skeletons,
and the fossil record. Fish. Bull. U.S. 70:565-584.
, AND R. J. Lavenberg. 1968. Deep-water fishes of California.
Univ. Calif Press, Berkeley.
,AND . 1975. Tidepool and nearshore fishes of California.
Fives, J. M. 1970a. BlenniidaeoftheNorth Atlantic. Cons. Int. Explor.
Mer, Fiches Identification Oeufs Poissons 3.
. 1970b. Investigations of the plankton of the west coast of
Ireland. IV. Larval and post-larval stages of fishes taken from the
plankton of the west coast in surveys during the years 1958-1966.
Proc. R. Ir. Acad. Sect. B 70:15-93.
Flegler, C. 1977. Electron microscopic studies on the development
of the chorion of the viviparous teleost Dennogenys pusillus (Hem-
irhamphidae). Cell Tissue Res. 179:255-270.
Foersch, W. 1966. Spawning Phractura ansorgei. Trop. Fish Hobby.
1966 (April):4-6, 8-9.
Follett, W. I. 1952. Annotated list offishes obtained by the California
Academy of Sciences during six cruises of the U.S.S. Mulberry
conducted by the U.S. Navy of Central California in 1949 and
1950. Proc. Calif Acad. Sci., Ser. 4, 27:399-432.
Ford, E. 1922. On the young stages of B/enn/;«oce//am(L.), Blennius
pholis{L.). and Blennius gallorugine(L.). J. Mar. Biol. Assoc. U.K.
12:688-692.
FoREV, P. L. 1973a. A revision of elopiform fishes, fossil and recent.
Bull. Br. Mus. (Nat. Hist.) Geol. Suppl. 10.
. 1973b. Relationships of elopomorphs. p. 351-368. //;. Inter-
relationships of fishes. P. H. Greenwood, R. S. Miles and C. Pat-
terson (eds.). Zool. J. Linn. Soc. 53(Suppl. 1.).
FoRSTER, R. P., AND S. K. HoNG. 1958. In vilro transport of dyes by
isolated renal tubules of the flounder as disclosed by direct visu-
alization: Intracellular accumulation and transcellular movement.
J. Cell. Comp. Physiol. 51:259-272.
Foster, N. R. 1967. Comparative studies on the biology of killifishes
(Pisces: Cyprinodontidae). Unpubl. PhD dissertation. Cornell Uni-
versity, NY.
. 1968. Utility of egg and larval characters in the classification
of atheriniform fishes. Am. Zool. 8:338.
. 1969. Field and aquarium observations on C/jr/opeoiWMpf «-
gelleyi Fowler, 1939. the Jamaican killifish. J. Am. Killifish Assoc.
6(1): 17-25.
. 1973. Behavior, development, and early life history of the
Asian needlefish, Xenenlodon cancila. Proc. Acad. Nat. Sci. Phila.
I25(5):77-88.
. 1974. Cyprinodontidae, p. 127-141. In: Manual for identi-
fication of early developmental stages offishes of the Potomac River
estuary. A. J. Lippson and R. L. Moran (eds.). Power Plant Siting
Program of Md. Dep. Nat. Resour. PPSP-MP-13.
. 1982. Literature review. J. Am. Killifish Assoc. 15(1 ):4-9.
, AND L. FuiMAN. MS. Ecology, behavior, and early life history
of the lyre goby, Evorthodus lyricus.
Folrmanoir,P. 1970. Notes ichthyologiques(l). Cah. ORSTOM Ser.
Oceanogr. 8:19-33.
. 1971a. Notes ichthyologiques (3). Cah. ORSTOM Ser. Ocean-
ogr. 9:267-278.
. 1971b. Notes ichthyologiques (4). Cah. ORSTOM Ser. Ocean-
ogr. 9:491-500.
. 1973. Notes ichthyologiques (5). Cah. ORSTOM Ser. Ocean-
ogr. 11:33-39.
. 1976. Formes post-larvaires et juveniles de poissons cotiers
pris au chalut pelagique dans le sud-ouest Pacifique. Cah. Pac. 19:
47-48.
Fowler, H. W. 1934. Descriptions of new fishes obtained 1907 to
1910, chiefly in the Philippine islands and adjacent seas. Proc. Acad.
Nat. Sci. Phila. 85:233-267.
. 1936. Carangidae, p. 675-756. /«; The marine fishes of West
Africa. Bull. Am. Mus. Nat. Hist. 70(2).
. 1944. The fishes, p. 57-583. In: Results of the Fifth George
Vanderbilt expedition (1941). Monogr. Acad. Nat. Sci. Phila. 6.
. 1945. A study of the fishes of the southern Piedmont and
Univ. Calif Press, Berkeley.
coastal plain. Monogr. Acad. Nat. Sci. Phila. 7.
Francis, M. P. 1981. Meristic and morphometric variation in the
lancetfish, Alepisaurus, with notes on the distribution of A. ferox
and A. breviroslris. N.Z. J. Zool. 8:403-408.
704
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Frank, K. T., and W. C. Legoett, 1983. Multispecies larval fish
associations: Accident or adaptation? Can. J. Fish. Aquat. Sci. 40:
754-762.
Frank, S. 1972. Entwicklung der Eier des Langflossensalmlers Bry-
conalestes longipinnis(GviiHheT. 1864)(Pisces: Characinoidei). Vestn.
Cesk. Spol. Zool. 36:237-240.
. 1978. Beobachtung der Embryonalentwicklung. Beispiel: das
Fischei. Mikrokosmos 1978:251-255.
Franke, H.-J. 1979. Algenbekampfung ohne Chemie, Zucht des al-
genfressenden Antennenwelses ,-l«a.s(n« spec. Aquarian Mag. 1975:
216-221.
. 1981. Cfe«o/i(aw.s Aiyeto (Valenciennes in Cuvier and Val-
enciennes, 1849). Aquarien Terrarien 28:345-348.
Fraser, T. H. 1971. The fish Dinolestes lewim with comments on its
osteology and relationship. Jpn. J. Ichthyol. 18:157-163.
. 1972a. Comparative osteology of the shallow water cardinal
fishes (PercifiDrmes: Apogonidae) wtih reference to the systematics
and evolution of the family. Rhodes Univ. J. L. B. Smith Inst.
Ichthyol. Bull. No. 34.
. 1972b. Some thoughts about the teleostean fish concept — the
Paracanthopterygii. Jpn. J. Ichthyol. 19:232-242.
, and M. M. Smith. 1974. An exterilium larval fish from South
Africa with comments on its classification. Copeia 1974:886-892.
Fraser-Brunner, A. 1931. SomeinterestingWest African fishes, with
descriptions of a new genus and two new species. Ann. Mag. Nat.
Hist., Ser. 8(45):2 17-225.
. 1949. A classification of the fishes of the family Myctophidae.
Proc. Zool. Soc. Lend. 1 18(4):1019-1 106.
. 1950. Studies in plectognath fishes from the "Dana" Expe-
ditions. I. An interesting new genus of triacanthodid fishes from
the Celebes Sea. Dana-Rep. Carlsberg Found. 35:1-8.
Freeman, H. W. 1951. Contribution to the evolution and classification
of the fishes of the family Agonidae. Unpubl. PhD dissertation.
Stanford Univ., Stanford, CA.
Freihofer, W. C. 1963. Patterns of the ramus lateralis accessorius
and their systematic significance in teleostean fishes. Stanford Ich-
thyol. Bull. 8:80-189.
. 1970. Some nerve patterns and their systematic significance
in paracanthopterygian, salmoniform, gobioid, and apogonid fishes.
Proc. Calif Acad. Sci., Ser. 4. 38:215-264.
. 1978. Cranial nerves of a percoid fish, Polycentrus schom-
hrugkii (family Nandidae), a contribution to the morphology and
classification of the order Perciformes. Occas. Pap. Calif Acad. Sci.
128.
Fricke, H. W. 1980. Mating systems, maternal and biparental care
in triggerfish (Balistidae). Z. Tierpsychol. 53(2):105-122.
Fricke, R. 1980. Neue Fundorte und noch nicht beschriebene Ge-
schlechtsunterschiede einiger Arten der Gattung Calllonynnis ( Pisces,
Perciformes, Callionymidae), mil Bemerkungen zur Systematik in-
nerhalb dieser Gattung und Beschreibung einer neuen Untergattung
und einer neuen Art. Ann. Mus. Civ. Stor. Nat. Giacomo Doria
83:57-105.
. 1981a. Revision of the genus Synchirnpus (Teleostei: Cal-
lionymidae). J. Cramer, Braunschweig, F.R.G.
. 1981b. The kaianus-gvouxt of the genus Callionymus (Pisces:
Callionymidae), with descriptions of six new species. Proc. Cal.
Acad. Sci. 42:349-377.
Fridgeirsson, E. 1978. Embryonic development of five species of
gadoid fishes in Icelandic waters. Rit Fiskideildar 5.
Fritzsche, R. a. 1978. Development of fishes of the Mid-Atlantic
Bight. An atlas of egg, larval and juvenile stages. Vol. V. Chaeto-
donitidae through Ophidiidae. U.S. Fish Wildl. Ser. Biol. Serv.
Prog. FWS/OBS-78/12.
. 1980. Revision of the eastern Pacific Syngnathidae (Pisces:
Syngnathiformes), including both recent and fossil forms. Proc.
Calif. Acad. Sci. 42(6):181-227.
. 1982. Osteichthyes, p. 858-944. In: Synopsis and classifica-
tion of living organisms. S. P. Parker (ed.). McGraw-Hill. Inc., New
\ork.
, AND G. D. Johnson. 1980. Early osteological development of
white perch and striped bass with emphasis on identification of
their larvae. Trans. Am. Fish. Soc. 109:387-406.
Frost, B. W., and L. E. McCrone. 1979. Vertical distribution, diel
vertical migration and abundance of some mesopelagic fishes in
the eastern subarctic Pacific Ocean. Fish. Bull. U.S. 76:751-770.
FiiiMAN, L. A. 1976. Notes on the early development of the sea raven,
Hernitriplerus amencanus. Fish. Bull. U.S. 74:467-470.
. 1979. Descriptions and comparisons of catoslomid fish larvae:
northern Atlantic drainage species. Trans. Am. Fish. Soc. 108:560-
603.
. 1982a. Correspondence of myomeres and vertebrae and their
natural variability during the first year of life in yellow perch, p.
56-69. In: Proceedings of the Fifth Annual Larval Fish Conference.
C. F. Bryan, J. V. Conner and F. M. Truesdale. (eds). Louisiana
Cooperative Fishery Research Unit, Louisiana State L'niversily,
Baton Rouge, LA.
. 1982b. Family Esocidae, pikes, p. 155-193. //(.■ Identification
of larval fishes of the Great Lakes Basin with emphasis on the Lake
Michigan drainage. N. A. Auer (ed.). Ann Arbor, Michigan, Great
Lakes Fish. Comm. Spec. Publ. 82-3.
. 1983a. Analysis of egg and larval characters for their use in
the systematics of ostariophysan fishes. Unpubl. PhD dissertation.
University of Michigan. Ann Arbor, Ml.
. 1983b. Growth gradients in fish larvae. J. Fish Biol. 23(1):
117-123.
, J. V. Conner, B. F. Lathrop, G. L. Buvnak, D. E. Snyder
AND J. J. Loos. 1983. Slate of the art of identification for cyprinid
fish larvae from eastern North America. Trans. Am. Fish. Soc.
112(2B):319-322.
FijiTA, K. 1971. Early development of the squamation in Titapia
sparrmanni. Jpn. J. Ichthyol. 18:90-93.
Fi^jiTA, S. 1955. On the development and prelarval stages of the file-
fish, Monacanlhus cirrhifer Temminck et Schlegel. Sci. Bull. Fac.
Agri. Kyushu Univ. 1 5(2):229-234. (In Japanese, English sum-
mary.)
. 1957a. On the development and prelarval stage of a damsel
fish, Chromis notatus (Temminck et Schlegel). Jpn. J. Ichthyol. 6:
87-90.
. 1957b. On the larval stages of a scorpaenid fish, Sebastes
pachycephalus nigricans (Schmidt). Jap. J. Ichthyol. 6:91-93. (In
Japanese. English summary.)
. 1958. On the egg development and larval stages of a viviparous
scorpaenid fish, Sebastes oblongns Gunther. Bull. Jpn. Soc. Sci.
Fish. 24:475-479. (In Japanese, English summary.)
. 1959. On the egg development of a viviparous scorpaenid fish
Sebastes pachycephalus nigricans (Schmidt). J. Jpn. Assoc. Zool.
Gardens and Aquariums l(2):42-43.
. 1960. Egg development and prelarval stages of a silver-belly,
Leiognathus nuchalis (Temminck et Schlegel). Bull. Jap. Soc. Sci.
Fish. 26:1091-1094.
. 1962. Studies on the life history of common pufferfishes in
Japan. Rep. Nagasaki Prefect. Fish. Res. Stn. 2. (In Japanese.)
. 1965. Early development and reanng of two common flat-
fishes, Eopsetta gngorjewi (Herzenstein) and Tanakius kitaharai
(Jordan et Starks). Bull. Jpn. Soc. Sci. Fish. 31:258-262.
. 1966. Egg development, larval stages and reanng of the puffer.
Lagocephalus hinaris spadiceus (Richardson). Jpn. J. Ichthyol. 13:
162-168. (In Japanese, English summary.)
, AND S. MiTO. 1960. Egg development and hatched larvae of
a chaetodontid fish, Chaelodontoplus septentnonalis (Temminck et
Schlegel). Bull. Jap. Soc. Sci. Fish. 26:227-29. (In Japanese, English
abstract.)
— , AND K. Nakahara. 1955. On the development of the egg and
the prelarval stages of a scorpaenid fish, Inimicusjapomcus (Cuvier
et Valenciennes). Sci. Bull. Fac. Agric. Kyushu Univ. 15:223-228.
(In Japanese, English summary.)
— , AND T. Takita. 1965. Egg development and prelarval stages
LITERATURE CITED
705
of a sole, Areliscus trigrammus (Gunthcr). Bull. Jpn. Soc. Sci. Fish.
31:488-492. (In Japanese, English abstract.)
— , AND K. UcHiDA. 1957. On the development of the egg and
prelarval stages of a sole, Cynoglossus rohuslus Gunther. Sci. Bull.
Fac. Agric. Kyushu Univ. 16:319-322. (In Japanese, English sum-
mary.)
— , AND M. Ueno. 1954. On the development of the egg and
prelarval stages of Siganiis fuscescens (Houttuyn) by artificial in-
semination. Jpn. J. Ichthyol. 3:129-132.
FuKL'HARA, O. 1971. Development of eggs and larvae of .-ljfra»;»w«
agrammus. Suisan Yoshoka 19:159-165. (In Japanese.)
, andK. Ito. 1978. On the formation of the fins and squamation
in the Japanese parrot fish, Opiegnathus fascialus reared in the
laboratory. Bull. Nansei Reg. Fish. Res. Lab. 1 1:9-17.
FnKUSHO, K. 1975. Development of color pattern in Opiegnalhus
fasciatus. useful indicator in rearing. Jpn. J. Ichthyol. 22:23-30.
Fulton, T. W. 1898. The ovaries and ovarian eggs of the angler or
frog-fish (Lophius piscalorius) and of the John Dory (Zeus faher).
Ann. Rep. Fish. Board Scotl. 16:125-134.
Fi'RiiK.AWA, I, 1951. Clupeoid fry possibly referable to Gonorynchiis
ahbrcviatus Temmmck et Schlegel. Jpn. J. Ichthyol. 4:292-294. (In
Japanese, English summary.)
FuSTER DE Plaza, M. L., and J. C. Plaza. 1955. Nuevos ensayos para
obtener la reproduccion artificial de las percas o truchas criollas
(Percichthys sp.). Misc. Publ. Minist. Agric. Ganad. (Buenos Aires)
407.
FuTrH. C. R. 1971. Ijarvae of Monolene sessilicauda Goodt, 1880
(Bothidae). Fla. Dep. Nat. Resour. Leafl. Ser. 4 (Pi. I, No. 21).
. 1977. Larvae of Tnchopsetta ventralis (Pisces: Bothidae), with
comments on intergeneric relationships within the Bothidae. Bull.
Mar. Sci. 27:740-757.
, AND F. H. HoFF, Jr. 1971. Larval development of Syacn/m
papillosum (Bothidae) with notes on adult morphology. Fla. Dep.
Nat. Resour. Leafl. Ser. 4 (Pt. I, No. 20).
, R. W. Topp andE. D. HoLiDE. 1972. Developmental osteology
of the lined sole, Achints lineatus (Pisces: Soleidae). Contrib. Mar.
Sci. 16:33-58.
Gabriel. M. L. 1944. Factors affecting the number and form of ver-
tebrae in Fundiitus heteroclUtis. J. Exp. Zool. 95:105-143.
Gamble, J. C, AND E. D. HouDE. 1984. Growth, mortality and feeding
of cod (Gadus morhua L.) larvae in enclosed water columns and
in laboratory tanks, p. 123-143. In: The propagation of cod. E.
Dahl, D. Danielssen, E. Moksness and P. Solemdal (eds.). Flode-
vigen Rapp. 1.
. P. Maclachlan.N.T. Nicoll AND I. G.Baxter. 1981. Growth
and feeding of Atlantic herring reared in large plastic enclosures.
Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:121-134.
Garman, S. 1892. The discoboli. Mem. Mus. Comp. Zool. Harv.
14(2): 1-96.
Garnaud, J. I960. La ponte, I'eclosion, la larve du baliste Balistes
caphscus Linne 1758. Bull. Inst. Oceanogr. Monaco 1081:1-6.
Garrett, G. P. 1982. Variation in the reproductive traits of the Pecos
pupfish, Cyprinodon pecosensis. Am. Midi. Nat. 180:355-363.
Gatesoupe, F.-J., AND P. LuQUET. 1981. Practical diet for mass culture
of the rotifer Brachionus plicatitis: application to larval rearing of
sea bass, Dicenlrarchus labrax. Aquaculture 22:149-163.
, AND J. Robin. 1981. Commercial single-cell proteins either
as sole food source or in formulated diets for intensive and con-
tinuous production of rotifers (Brachionus plicalilis). Aquaculture
25:1-15.
Galildie, R. W., E. J. Gravnoth and J. Illinc^worth. 1980. The
relationship of the iron content of some fish otoliths to temperature.
Comp. Biochem. Physiol. 66:19-24.
Geffen, A. 1982. Otolith nng deposition in relation to growth rate in
hemng (Clupea harengus) and turbot (Scophlhalmus maximus)
larvae. Mar. Biol. 71:317-326.
Gehringer, J. W. 1959a. Early development and metamorphosis of
the ten-pounder £/opiiaHrwi Linnaeus. U.S. Fish Wildl. Serv. Fish.
Bull. 59(155):619-647.
. 1959b. Lcptocephalus of the Atlantic tarpon, Megalops al-
lanticus Valenciennes, from offshore waters. Q. J. Fla. Acad. Sci.
2I(3):235-240.
George, A., and V. G. Springer. 1980. Revision of the clinid fish
tribe Ophiclinini, including five new species, and definition of the
family Clinidae. Smithson. Contrib. Zool. 307:1-31.
Gerking, S. D., and R. R. Ralisch. 1979. Relative importance of
size and chronological age in the life programme of fishes. Arch.
Hydrobiol. 13:181-194.
Gery. J. 1977. Characoids of the World. T.F.H. Publications, Inc.,
Neptune City, NJ.
GiBBS, R. H., Jr. 1959. Asynopsisofthepost larva of western Atlantic
lizard-fishes (Synodontidae). Copeia 1959:232-236.
. 1964a. Family Astronesthidae, p. 1-10. /ai. Fishes of the west-
em North Atlantic. H. B. Bigelow (ed.). Mem. Sears Found. Mar.
Res. No. 1 (Pt. 4).
. 1964b. Family Idiacanthidae, p. 512-522. In: Fishes of the
Western North Atlantic. H. B. Bigelow (ed.). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 4).
, T. A. Clarke and J. R. Gomon. 1983. Taxonomy and dis-
tribution of the stomioid fish Genus Eustomias (Melanostomiidae),
I: Subgenus Nominostomias. Smithson. Contrib. Zool. No. 380.
, and B. B. Collette. 1 959. On the identification, distribution,
and biology of the dolphins Coryphaena hippunis and C. equiselis.
Bull. Mar. Sci. GulfCanbb. 9:117-152.
, and . 1967. Comparative anatomy and systematics of
the tunas, genus Thunnus. U.S. Fish. Wildl. Serv. Fish. Bull. 66:
65-130.
-. and J. C. Staiger. 1970. Eastern tropical Atlantic flyingfishes
of the genus Cypseturus (Exocoetidae). Stud. Trop. Oceanogr.
(Miami) No. 4(Pt. 2):432-466.
Gilbert, C. H. 1893. Scientific results of explorations by the U.S.
Fish Commission Steamer Albatross. No. XXII. Descriptions of
thirty-four new species of fishes collected in 1888 and 1889, prin-
cipally among the Santa Barbara Islands and in the Gulf of Cali-
fornia': Proc. U.S. Natl. Mus. l4(880):539-566.
. 1905. The deep sea fishes of the Hawaiian Islands. Bull. U.S.
Fish Comm. 23(2):577-713.
, AND C. V. Burke. 1912. Fishes from Bering Sea and Kam-
chatka. Bull. Bur. Fish. Wash. 30(19101:31-96.
, AND C. L. HuBBS. 1916. Report on the Japanese macrourid
fishes collected by the United States fisheries steamer "Albatross"
in 1906, with a synopsis of the genera. Proc. U.S. Natl. Mus. 51(2149):
135-214.
, AND W. F. Thompson. 1916. Family Macrouridae, p. 471-
476. In: Fishes collected by the United States Bureau of Fisheries
steamer "Albatross" during 1888, between Montevideo, Uruguay,
and Tome, Chile, on the voyage through the Straits of Magellan.
W. F. Thompson (ed.). Proc. U.S. Natl. Mus. 50(2133):40l-476.
Gilbert, C. R. 1980. Umbra limi. Umbra pygmaea. p. 129-130. In:
Atlas of North Amencan freshwater fishes. D. S. Lee, C. R. Gilbert,
C. H. Hocutt, R. E. Jenkins, D. E. McAllister and J. R. StaufTer,
Jr. (eds.). North Carolina Mus. Nat. Hist., Raleigh, NC.
Gilchrist, J. D. F. 1903. The development of South -African fishes.
Part 1. Mar. Invest. S. Afr. 2:181-201.
. 1904. The development of South African fishes. Part 2. Mar.
Invest. S. Afr. 3:131-150.
. 1918. The eggs and spawning-habits ofthe pilot fish (.Vawcrarw
ductor). Ann. Mag. Nat. Hist., Ser. 9, 2(7): 1 14-1 18.
-. AND H. Hunter. 1919. Reproduction of fishes in Table Bay.
Trans. R. Soc. S. Afr. 8(1): 1-1 6.
Gill, T. N. 1862. Synopsis of the polynematoids. Proc. Acad. Nat.
Sci. Phila. 13:271-282.
. 1863. Descriptive enumeration of a collection of fishes from
the western coast of Central America. Proc. Acad. Nat. Sci. Phila.
15(4): 153-200.
. 1865. Synopsis ofthecyclopteroidsofeastem North America.
Proc. Acad. Nat. Sci. Phila. 16:189-194.
706
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
— . 1872. Arrangement of the families of fishes, or classes Pisces,
Marsipobranchii and Leptocardii. Smilhson. Misc. Collect. 1 1:247.
— . 1888. On the classification of the mail-cheeked fishes. Proc.
U.S. Natl. Mus. 11:567-592.
— . 1891. On the relations of Cyclopteroidea. Proc. U.S. Natl.
Mus. 13:361-376.
— . 1893. A comparison of antipodal faunas. Mem. Natl. Acad.
Sci. 6:91-124.
GiNSBURG. I. 1952. Fishes of the family Carangidae of the northern
Gulf of Mexico and three related species. Publ. Inst. Mar. Sci. Univ.
Tex. 2:43-117.
. 1955. Fishes of the family Percophididae from the coasts of
eastern United States and the West Indies, with descriptions of four
new species. Proc. U.S. Natl. Mus. l04(3347):623-649.
GiNZBURG, A. S. 1972. Fertilization in fishes and the problem of
polyspermy. Acad. Sci. USSR, Inst. Develop. Biol. 1968. Trans-
lated from Russian by the Isr. Program Sci. Transl.
GJ0SAETER, J. 1973. Age, growth and mortality of the myctophid fish,
Benthosema glaciale {Reinhardi), from western Norway. Sarsia 52:
1-14.
, AND K. Kawaguchi. 1980. A review of the world resources
of mesopelagic fish. FAO Fish. Tech. Pap. 193.
GoiN, C. J., AND D. S. Erdman. 195 1 . The crested oarfish, Lophotus
lacepedei. from Florida; first record from the western North Atlan-
tic. Copeia 1951:285-288.
Gold, J. R. 1979. Cytogenetics, p. 353-405. //;. Fish Physiology. W.
S. Hoar, D. J. Randall and J. R. Brett (eds.). Academic Press, New
York.
Golvan, Y.-J. 1962. Catalogue systematique des noms de genres de
poissons actuels. Ann. Parasitol. Hum. Comp. 37(6): 1-227.
Gomez Gaspar, A. 1981. Observaciones taxonomicas y ecologicas de
estados larvales Elopiformes en la Laguna de la Restinga y areas
proximas (Isla Margarita, Venezuela). Bol. Mus. Mar. Univ. Bogota
10:3-23.
Gonzales, T. 1946. Contribute alia conoscenza dello sviluppo post-
embrionale in Microichlhys coccoi. Ruppell. Boll. Pesca Pisci. Id-
robiol. 1:39-46.
GooDE, G. B., AND T.H. Bean. 1896. Oceanic Ichthyology. U.S. Natl.
Mus. Spec. Bull.
Goody, P. C. 1969. The relationships of certain Upper Crealaceous
teleosts with special reference to the myctophoids. Bull. Bnt. Mus.
(Nat. Hist.) Geol. Suppl. 7.
Goodyear, R. H. 1970. A new species of .-l^a-vo/epw, a bathypelagic
fish from the Gulf of Panama (Pisces, Lampridiformes, Megalo-
mycteridae). Steenstrupia 1(3): 17-20.
GopiNATH, K. 1946. Notes on the larval and post-larval stages of fishes
found along the Trivandrum coast. Proc. Indian Natl. Sci. Acad.
12(l):7-21.
GoRBUNOVA, N. N. 1954. The reproduction and development of the
walleye pollock, Theragra chalcogramma (Pallas). Tr. Inst. Okean-
ol. Akad. Nauk SSSR 11:132-195.
. 1 964a. Breeding and development of hemilepidotine sculpins
(Cottidae, Pisces), p. 249-266. In: Fishes of the Pacific and Indian
Oceans. Biology and distribution. T. S. Rass (ed.). (Isr. Program
Sci. Transl. No. 1411.)
. 1 964b. Spawning and development of greenlings (family Hex-
agrammidae), p. 121-185. In: Greenlings: taxonomy, biology, in-
teroceanic transplantation. T. S. Rass (ed.). Springfield, Va. (Transl.
available from: NTIS, TT69-55097.)
. 1981. Larvae of the genus (7Hf/^/(>rna(Gonostomatidae) with
keys. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 21(4):I38-141.
. 1982. Larvae of trichiuroid fishes from collection of the In-
ternational Mexican Biological Center (Gempylidac, Trichiuridae,
Pisces). Proc. P. P. Shirshov Inst. Oceanol. 1 18:85-104. (In Rus-
sian.)
— , AND N. V. Parin. 1963. Development of eggs and larvae of
the flying fish, Cheilopogon unicolor (Cuv. et Val.). Tr. Inst. Okean-
ol. Akad. Nauk SSSR 62:62-67. (In Russian, English abstract.)
, AND O. A. Zvyagina. 1975. Eggs and larvae of the scaled
sardine (Harengiila pensacolae) (Pisces, Clupeidae). J. Ichthyol.
(Engl. Transl. Vopr. Ikhtiol.) 15:829-833.
Gordon, D. J. 1982. Systematics and distribution of larval fishes of
the subfamily Ophidiinae (Pisces, Ophidiidae) in the eastern Gulf
of Mexico. Unpubl. MS thesis. Univ. Miami, Coral Gables, FL.
GosLiNE, W. A. 1948. Some possible uses of X-rays in ichthyology
and fishery research. Copeia 1948:58-61.
. 1951. The osteology and classification of the ophichthid eels
of the Hawaiian Islands. Pac. Sci. 5:298-320.
. 1955. Theosteology and relationships of certain gobioid fishes,
with particular reference to the genera Kraemaria and Microdes-
mus. Pac. Sci. 9:158-170.
. 1960a. Contributions toward a classification of modem iso-
spondylous fishes. Bull. Br. Mus. Nat. Hist. Zool. 6:325-365.
. 1 960b. Hawaiian lava-flow fishes. Part IV. Snyderidia canina
Gilberi, with notes on the osteology of ophidioid families. Pac. Sci.
14:373-381.
. 1961. Some osteological features of modem lower teleostean
fishes. Smithson. Misc. Collect. 142:1-42.
. 1 962. Systematic position and relationships of the percesocine
fishes. Pac. Sci. 16:207-217.
. 1963a. Considerations regarding the relationships of the Per-
copsiform, Cyprinodontiform, and Gadiform fishes. Occas. Pap.
Mus. Zool. Univ. Mich. No. 629.
. 1963b. Notes on the osteology and systematic position of
Hypoptychus dybowskii Steindachner and other elongate perciform
fishes. Pac. Sci. 17:90-101.
. 1966. The limits of the fish family Serranidae, with notes on
other lower percoids. Proc. Calif Acad. Sci. 33:91-1 1 1.
— . 1968. The suborders of perciform fishes. Proc. U.S. Natl. Mus.
124(3647): 1-78.
— . 1970. A reinterpretation of the teleostean fish order Gobie-
sociformes. Proc. Cal. Acad. Sci. 38:363-382.
. 1971. Functional morphology and classification of teleostean
fishes. Univ. Press Hawaii, Honolulu.
— . 1978. Unbranched dorsal-fin rays and subfamily classification
in the fish family Cyprinidae. Occas. Pap. Mus. Zool. Univ. Mich.
684.
. 1980. The evolution of some structural systems with reference
to the interrelationships of modem lower Teleostean fish groups.
Jpn. J. Ichthyol. 27:1-28.
, N. B. Marshall and G. W. Mead. 1966. Order Iniomi:
characters and synopsis of the families. Mem. Sears Found. Mar.
Res. 21(Pt. 5):l-29.
GoszTON-i I, A. E. 1977. Results of the research cruises of FRV "Wal-
ther Herwig" to South America. XLVIII. Revision of the South
American Zoarcidae (Osteichthyes. Blennioidei), with the descrip-
tion of three new genera and five new species. Arch. Fischereiwiss.
27:191-249.
GoTSHALL, D. W. 1971. Occurrences of the bearded eelpout, Lyco-
nema barbatum. off northern Califomia and southem Oregon. Cal-
if Fish Game 57:295-297.
Gould, W. R. 1965. The biology and morphology oi Acyrtops ber-
ylUnus. the emerald clingfish. Bull. Mar. Sci. Gulf Canbb. 15:165-
188.
Gould, S. J. 1977. Ontogeny and phylogeny. Belknap Press, Cam-
bridge, MA.
GovoNi.J. J. 1980. Morphological, histological and functional aspects
of alimentary canal and associated organ development in larval
Leiostomus xanthurus. Rev. Can. Biol. 39:69-80.
, J. E. Olnev. D. F. Markle and W. R. Curtsinger. 1984.
Observations on stracture and evaluation of possible functions of
the vexillum in larval Carapidae (Ophidiiformes). Bull. Mar. Sci.
34:60-70.
Graham, D. H. 1939. Breeding habits of fishes of Otago Harbour and
adjacent seas. Trans. R. Soc. N.Z. 69:361-372.
Grassi, G. B. 1913. Metamorfosi dei murenoidi. Ricerche sistema-
tische ed ecologische. Fischer, Jena.
, and S. a. Calandruccio. 1897. Descrizione d"un Lepra-
LITERATURE CITED
707
cephalus breviroslris in via tranformarzi in Anguilta vulgaris. Atti
Accad. Naz. Lincei, Ser. 6. 5(Pt. l):239-240.
Graves, J. E., and G. N. Somero. 1982. Electrophoretic and func-
tional enzymic evolution in four species of eastern Pacific barra-
cudas from different thermal environments. Evolution 36:97-106.
Greenfield, D.W. 1965. Systematics and zoogeography of .V/vnpm«i
in the eastern tropical Pacific. Calif Fish Game 51:229-247.
. 1972. Notes on the biology of the arrow blenny L!<cav'aWe«-
nius zingaro (Bohlke) from Bntish Honduras. Copeia 1972:591-
592.
-. 1974. A revision of the squirrelfish genus AAnprH/ii Cuvier
(Pisces: Holocentridae). Nat. Hist. Mus. Los Ang. Cty. Sci. Bull.
19.
Greenwood, P. H. 1955. Reproduction in the catfish, C/ariai wo5-
sambicus Peters. Nature (Lond.) 176:516-519.
. 1956. The reproduction of Clarias mossambicus Peters in
Uke Victoria. Sci. Counc. Afr. South Sahara Publ. 25:77-78.
. 1973. Interrelationships of osteoglossomorphs, p. 307-332.
In: P. H. Greenwood. R. S. Miles and C. Patterson (eds.). Inter-
relationships of fishes. Academic Press, New York.
. 1976. A review of the family Centropomidae (Pisces, Perci-
formes). Bull. Bnt. Mus. (Nat. Hist.) Zool. 29:1-81.
. 1977. Notes on the anatomy and classification of elopomorph
fishes. Bull. Br. Mus. (Nat. Hist.) Zool. 32(4):65-102.
,G.S. Myers, D.E.Rosen AND S. H.Weitzman. 1967. Named
main divisions of teleostean fishes. Proc. Biol. Soc. Wash. 80:227-
228.
, AND D. E. Rosen. 1971. Notes on the structure and relation-
ships of the alepocephaloid fishes. Am. Mus. Novit. 2473.
-, S. H. Weitzman and G. S. Myers. 1966. Phyletic
studies of teleostean fishes, with a provisional classification of living
forms. Bull. Amer. Mus. Nat. Hist. 131:339-456.
Gregory, W. K. 1933. Fish skulls. A study of the evolution of natural
mechanisms. Trans. Am. Philos. Soc. 23:75-481.
. 1959. Fish skulls. A study of the evolution of natural mech-
anisms. Noble Offset Printers, Inc., New York. (First published in
1933. Trans. Am. Philos. Soc. 23:75^81.)
, and G. M. Conrad. 1936. Pictorial phylogenies of deep-sea
isopondyli and iniomi. Copeia 1936:21-36.
,AND . 1937. The comparative osteology of the swordfish
(Xiphias) and the sailfish (Istiophorm). Am. Mus. Novit. 952.
, AND . 1943. The osteology of Liivarus impenalis. a
scombroid fish: a study in adaptive evolution. Bull. Am. Mus. Nat.
Hist. 31(2):225-283.
Grey, M. 1953. Fishes of the family Gempylidae, with records of
Nesiarchus and Epirmula from the western Atlantic and descrip-
tions of two new subspecies of Epwnula onenlalis. Copeia 1953:
135-141.
. 1955a. Notes on a collection of Bermuda deepsea fishes. Field-
iana Zool. 37:265-302.
. 1955b. The fishes of the genus Telragonurus Risso. Dana-
Rep. Carlsberg Found. 4 1 .
. 1 960. Description of a western Atlantic specimen of Scom-
brolabrax helerolepis Roule and notes on fishes of the family Gem-
pylidae, Copeia 1960:210-215.
1964. Family Gonostomatidae, p. 78-240. In: Fishes of the
western North Atlantic. H. B. Bigelow (ed). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 4).
Grioorev, G. v., AND V. P. Serebri AKOv. 1981. Eggs of roundnose
grenadier, Coryphaenoides rupestns Gunnerus 1765. J. Northwest
Atl. Fish. Sci. 2:73-74.
Grimes, P. J., and D. A. Robertson. 1981. Egg and larval develop-
ment of the silver warehou, Senolella punclala (Pisces: Centrolo-
phidae). N.Z. J. Mar. Freshwater Res. 15:261-266.
Grimm, H. 1979a. Egg size and fecundity within the genus .-Ip/wn/iw
(Pisces, Cyprinodontidae). Meersforsch. Rep. Mar. Res. 27(3): 177-
185.
. 1979b. Investigations of the evolution of differently scaled
populations found in the species Aphanius analoliac (Leidenfrost,
1912) (Pisces, Cyprinodontidae). Genetics of the egg size. Z. Zool.
Syst. Evolutionsforsch. 17(l):66-77.
Grindley, J. R.. AND M. J. Penrith. 1965. Notes on the bathypelagic
fauna of the seas around South Africa. Zool. Afr. l(2):275-295.
GiiDOER, E. W. 1905. The breeding habits and the segmentation of
the egg of the pipefish {Siphosloma Jloridae). Proc. U.S. Natl. Mus.
29(1431):447-500.
. 1912. Oral gestation in the gaff-topsail cat-fish, Felichthysfeiis.
Science 35:192.
. 1916. The gaff-topsail (Felichthys felts): A sea catfish that
carries its eggs in its mouth. Zoologica (N.Y.) 2(5): 125-1 58.
. 1918. Oral gestation in the gaff-topsail catfish, Felichthysfeiis.
Carnegie Inst. Wash. Publ. 252:25-52.
. 1926. A study of the smallest shark-suckers (Echeneidae) on
record, with special reference to metamorphosis. Am. Mus. Novit.
234:1-26.
. 1928. The smallest known specimens of the sucking-fishes,
Reinora brachyptera and Rhombochirus osleochir. Am. Mus. Nov-
it. 294:1-5.
. 1930. The opah or moonfish, Lamprw /!<«a. on the coasts of
California and of Hawaii. Am. Nat. 65:531-541.
1931. The tnple-tail, Lobotes surinamensis. its names, oc-
currence on our coasts and its natural history. Am. Nat. 65:49-69.
Gltllard. R. R. L. 1975. Culture of phytoplankton for feeding marine
invertebrates, p. 29-60. /«. Culture of marine invertebrate animals.
W. L. Smith and M. H. Chanley (eds.). Plenum Publ. Corp., N.Y.
Glitart Manday, D., and D. Juarez Fernandez. 1966. Desarrollo
embrionario primeros estadios larvales de chema criolla, Epinephe-
lus stnalus (Bloch) (Perciformes: Serranidae). Acad. Cienc. Cuba
Estud.-Inst. Oceanol. 1:35-45.
Guitel, R. 1888. Recherches sur les Ledapogasters. Arch. Zool. Exp.
Gen., Ser. 2. 6:423-647.
GiiNTER, G., L. L. Sulya and B. E. Box. 1961. Some evolutionary
patterns in fishes' blood. Biol. Bull. (Woods Hole) 121:302-306.
GOnther, A. 1861. Catalogue of the Acanthopterygian fishes in the
collection of the British Museum. Vol. 3^. Bntish Museum, Lon-
don.
. 1862. Chromides. p. 264-316. //;. Catalogue of the fishes in
the British Museum. Vol. 4.
. 1887. Report on the deep-sea fishes collected by H. M.S. Chal-
lenger dunng the years 1873-76. Rep. Sci. Results Challenger, Zool.
22:1-335. (Repnnt 1963. Wheldon and Wesley, Ltd., and Hafner
Publishing Co., New York.)
. 1889. Report on the pelagic fishes collected by H. M.S. Chal-
lenger during the years 1873-1876. Rep. Sci. Results Challenger,
Zool. 31:1-44.
Gustato, G. 1976. Osservazioni suUa biologia e sul comportamento
di Carapus acus (Ophidioidea, Percomorphi). Boll. Soc. Nat. Na-
poli 85:505-535.
GuTHERZ, E. J. 1970. Characteristics of some larval bothid flatfish.
and development and distribution of larval spotfin flounder, Cy-
dopsetta fimbnata (Bothidae). U.S. Fish. Wild. Serv. Fish. Bull.
68:261-283.
Haak£, P. W., C. A. Wilson and J. M. Dean. 1982. A technique for
the examination of otoliths by SEM with application to lars'al fishes.
In: Proc. 5th Annual Larval Fish Conference, Baton Rouge, LA.
H^EDRiCH, R. L. 1967. The stromateoid fishes: systematics and a
classification. Bull. Mus. Comp. Zool. Harv. Univ. 135:31-139.
. 1969. A new family of stromateoid fishes from the equatorial
Indo-Pacific. Dana-Rep. Carlsberg Found. 76.
. 1970. A new species in the nomeid fish genus Psenes from
the equatorial eastern Pacific. Breviora 351.
. 1971. Theponsmoultoni:asignificantcharacter. Copeia 1971:
167-169.
. 1974. A rotating eye in rraf/i/p/pnYi. Copeia 1974:972-973.
, and M. H. Horn. 1972. A key to the stromateoid fishes.
Woods Hole Oceanogr. Inst. Tech. Rep. 72-15.
, and J. G. Nielsen. 1966. Fishes eaten by .-l/ep/iai/riu (Pisces,
708
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Iniomi) in the southeastern Pacific Ocean. Deep-Sea Res. 13:909-
919.
Haegele, C. W., AND J. F. ScHWEiGERT. 1984. Distribution and char-
acteristics of herring spawning grounds and description of spawning
behavior. Can. J. Spec. Publ. Fish Aquat. Sci. In press.
Haigh, E. H. 1972a. Larval development of three species of econom-
ically important South African fishes. Ann. S. Afr. Mus. 59(3):47-
70.
. 1972b. Development of Trachurus Irac/ninis (Carangidae),
the south African maasbanker. Ann. S. Afr. Mus. 59(8): 139- 149.
Hallacher, L. E. 1974. The comparative morphology of extrinsic
gasbladder musculature in the scorpionfish genus Sehastes (Pisces:
Scorpaenidae). Proc. Calif Acad. Sci. 40:59-86.
Halstead, B. W., D. D. Danielson. W. J. Baldwin and P. C. Engen.
1972. Morphology of the venom apparatus of the leatherback fish
Scomberoides sanclipeln (Cuvier). Toxicon 10:249-258.
Hamanda, K. 1961. Taxonomic and ecological studies of the genus
Hypomesus of Japan. Mem. Fac. Fish. Hokkaido Univ. 9(1): 1-56.
Hansen, E., AND P. J. Herring. 1977. Dual bioluminescent systems
in the anglerfish genus Ltnophryne. J. Zool. (Lond.) 182:103-124.
Hara, S., J. P. Canto, Jr., and J. M, E. Almendras. 1982. A com-
parative study of various extenders for milkfish, Chanos chanos
(Forsskal), sperm preservation. Aquaculture 28:339-346.
Harada, T. 1962. A contribution to the biology of the black rockfish,
Sehastes inermisCu\\er ex Valenciennes. Publ. Seto Mar. Biol. Lab.
10:307-361.
, K. MiZlJNO, O. MURATA, S. MlVASHITA AND H. HURLITANI.
1971. On the artificial fertilization and rearing of larvae in yel-
lowfin tuna. Mem. Fac. Agnc. Kinki Univ. (4):1 45-151.
, O. MuRATA AND H. FuRHTANi. 1973. On the artificial fertil-
ization and rearing of larvae in Marusoda, Auxis tapeinosoma.
Mem. Fac. Agric. Kinki Univ. (6):1 13-1 16.
, ANt5 S. MlVASHITA. 1973. On the artificial fertilization
and rearing of larvae in Hirasoda, Auxis Ihazard. Mem. Fac. Agric.
Kinki Univ. (6): 109-1 12.
, AND . 1974. On the artificial fertilization and
rearing of larvae in bonito. Mem. Fac. Agric. Kinki Univ. (7):l-4.
Harbison, G. R., L. P. Madin and N. R. SwANBERCi. 1978. On the
natural history and distribution of oceanic ctenophores. Deep-Sea
Res. 25:233-256.
Hardy, G. S. 1982. A revision of the fishes of the family Pentacer-
otidae (Perciformes). N.Z. J. Zool. 10:177-220.
Hardy, J. D., Jr. 1978a. Development of fishes of mid-Atlantic bight,
an atlas of egg, larval, and juvenile stages. Vol. II. Anguillidae
through Syngnathidae. U.S. Fish Wildl. Serv. Biol. Serv. Program
FWS/OBS-78/12.
. 1978b. Development of fishes of the mid-Atlantic Bight. An
atlas of egg, larval and juvenile stages. Vol. III. Aphredoderidae
through Rachycentridae. U.S. Fish Wildl. Ser. Biol. Serv. Program
FWS/OBS-78/12.
Harrington, R. W., Jr. 1958. Morphometry and ecology of small
tarpon Megalops allanlica Valenciennes from transitional stage
through onset of scale formation. Copeia 1958:1-10.
. 1959. Delayed hatching in stranded eggs of marsh killifish,
Fundutus confluenlus. Ecology 40:430-437.
. 1961. Oviparous hermaphroditic fish with internal self-fer-
tilization. Science 135(3492): 1749-1 750.
, AND J. S. Haeger. 1958. Prolonged natural deferment of
hatching in killifish. Science 1 28(3337):151 1.
Harrison, C. M. H., and G. Palmer. 1968. On the neotype of Ra-
diicephahis elongalus Osorio with remarks on its biology. Bull. Bnt.
Mus. (Nat. Hist.) Zool. 16:187-211.
Harry, R. R. 1 95 1. A new cusk-eel of the genus Ophidian from
California with notes on the genus. Stanford Ichthyol. Bull. 4:30-
35.
Hart.J.L. 1973. Pacific fishes of Canada. Bull. Fish. Res. Board Can.
180.
Hart, N. H., R. Pietri AND M. Donovan. 1984. The structure of the
chorion and associated surface filaments in On'r/ai— evidence for
the presence of extracellular tubules. J. Exp. Zool. 230:273-296.
Haryli, T., AND T. Nishiyama. 1981. Larval form of Zaprorid fish,
Zaprora silenus from the Bering Sea and the Northern Pacific. Jpn.
J. Ichthyol. 28:313-318.
Hastings, R. W., AND R. W. Yerger. 1971. Ecology and life history
of the diamond killifish, Adinia xenica (Jordan and Gilbert). Am.
Midi. Nat. 86:276-291.
H ATTORi, S. 1 964. Studies on fish larvae in the Kuroshio and adjacent
waters. Bull. Tokai Reg. Fish. Res. Lab. (40): 158 pp.
. 1970. Reproductive aspects of fish resources, p. 209-222. /?!.
Ocean development. 4. Exploitation of fisheries resources. T. Sasaki
(ed.). Ocean Dev. Cent. Press, Tokyo. (In Japanese.)
Hay, D. E. 1981. Effects of capture and fixation on gut contents and
body size of Pacific herring larvae. Rapp. P. V. Reun. Cons. Int.
Explor. Mer 178:395-400.
. 1984. Reproductive biology of Pacific herring. Can. J. Spec.
Publ. Fish Aquat. Sci. In Press.
Ha,YDOCK, I. 1971. Gonad maturation and hormone-induced spawn-
ing of the Gulf croacker, Bairdwila icistia. Fish. Bull. U.S. 69: 1 57-
180.
Hearne, M. 1983. Identification of larval and juvenile smelts (Os-
meridae) from Oregon and California using selected morphomctric
characters. Unpubl. MS thesis. San Francisco State University, San
Francisco, CA.
Hecht, T. 1978. A descriptive systematic study of the otoliths of the
neopterygian marine fishes of South Africa. Part I. Introduction.
Trans. R. Soc. S. Afr. 43(2): 191.
Heck, K. L., and M. P. Weinstein. 1978. Mimetic relationships
between tropical burrfishes and opisthobranchs. Biotropica 10:78-
79.
Heemstra, P. C. 1974. On the identity of certain Pacific and Caribbean
post-larval fishes (Perciformes) described by Henry Fowler. Proc.
Acad. Nat. Sci. Phila. 126:21-26.
. 1980. A revision of the zeid fishes (Zeiformes: Zeidae) of South
Africa. Rhodes Univ. J. L. B. Smith Inst. Ichthyol. Bull. 41.
. In press. Families Regalecidae, Lopholidae, Veliferidae. Lam-
pridae. In: Sea Fishes of South Africa, Grahamstown. S.A.
, andS. X. Kannemeyer. In press. The families Trachipteridae
and Radiicephalidae (Pisces: Lampriformcs) and a new species of
Zu from South Africa. Ann. S. Afr. Mus.
, and J. E. Randall. 1977. A revision of the Emmelichthvidae
(Pisces: Perciformes). Aust. J. Mar. Freshwater Res. 1977(28):36l-
396.
Hefford, a. E. 1910. Notes on teleostean ova and larvae observed
at Plymouth in spring and summer, 1 909. J. Mar. Biol. Assoc. U.K.
9:2-58.
Heincke, F., AND E. Ehrenbaum. 1900. Eier und Larven von Fischen
der Deutschen Bucht. II. Die Bestimmung der schwimmenden
Fischeier und die Methodik der Eimessungen. Wiss. Meeresuntcrs.
(Helgol.) 3:239-243.
Helfrk H, P. H. 1958. Theearly life history and reproductive behavior
of the maomao, Ahudefduf ahdoniinalis (Quoy and Gainard). Lln-
publ. PhD dissertation. Univ. Hawaii, Honolulu.
Hempel, G. 1979. Early life history of marine fish: the egg stage. Univ.
Washington Press, Seattle.
, and J. H. S. Blaxter. 1961. The experimental modification
of meristic characters in herring (.Clupea harcngits L.). J. Cons. Int.
Explor. Mer 26:336-346.
, AND H. Weikert. 1972. The neuston of the subtropical and
boreal Northeastern Atlantic Ocean. A Review. Mar. Biol. (Berl.)
13:70-88.
Hennig, W. 1950. Grundziige einer Theorie der phylogenetischen
Systematik. Deutscher Zentralverlag, Berlin.
. 1966. Phylogenetic Systematics. Univ. Illinois Press, Urbana.
Hensley, D. A. 1976. Collection of postlarval juvenile //op//ai ma/-
aharicus (Characoidei: Erythrinidae) in Florida. Fla. Sci. 39:236-
238.
LITERATURE CITED
709
. 1977, Larval development of Engyophn's senta (Bothidae),
with comments on intermuscular bones m flatfishes. Bull. Mar. Sci.
27:681-703.
Henzelmann, E. 1930. Exvjasuhex Jordanctlatlondue Goods el Bean.
(Floridakarpfling). Wochenschr. Aquar. Terranien. 27( 19):307-309.
Herald, E. S. 1939. The opah (La/xpra re^/«i) and its occurrence off
the California coast. Calif Fish Game 25:228-232.
Herre, A. W. 1942. New and little known phallostelhids, with keys
to the genera and Phillipine species. Stanford Ichthyol. Bull. 2: 1 37-
156.
Herring, P. J., and J. G. Morin. 1978. Bioluminescence in fishes,
p. 273-329. In: Bioluminescence in action. P. J. Herring (ed.).
Academic Press, London.
Herwig, N., and D. Dewev. 1982. Congrogadus subduscens: A new
challenge for the manne aquarist. Freshwater Mar. Aquarium 5:6-
91.
Hettler, W. 1981. Spawning and rearing Atlantic menhaden. Prog.
Fish Cult. 43:80-84.
Heufelder, G. R. 1982. Family Cottidae, sculpins, pp. 656-676. /«.-
Identification of larval fishes of the Great Lakes Basin with em-
phasis on the Lake Michigan drainage. N. A. Auer (ed.). Great
Lakes Fish. Comm. Spec. Publ. 82-3.
He> wood, V. H. (ED.). 1971. Scanning electron microscopy; system-
atic and evolutionary applications. Academic Press. New York.
Hkkei, G. M. 1979. Survival of fish larvae after injury. J. Exp. Mar.
Biol. Ecol. 37:1-17.
. 1982. Wound healing in fish larvae. J. Exp. Mar. Biol. Ecol.
57:149-168.
Hk K.LING, C. F. 1931. The structure of the otolith of the hake. Q. J.
Microsc. Sci. 74:547-561.
Hickman, C. P., Jr. 1959. The larval development of the sand sole
(Psettichthys melanosliclus). Wash. Dep. Fish. Res. Pap. 2(2):38-
47.
HiKiTA, T. 1 952. On the development o( Limanda schreuki (Schmidt).
Sci. Rep. Hokkaido Fish Hatchery 7( 1,2):133-144.
. 1958. On the development of long finned smelt. ,S'p;/-/«c7n«
lanceolalus. Sci. Rep. Hokkaido Fish Hatch. 13:1-4.
HiLDEBRAND, S. F. 1922. Notcs on habits and development of eggs
and larvae of the silversides, Memdia menidia and Menidia ber-
yllina. Bull. Bur. Fish. Wash. 38:1 13-120.
. 1946. A descriptive catalogue of the shore fishes of Peru. U.S.
Natl. Mus. Bull. 189.
. 1963a. Family Elopidae, p. 1 1 1-1 13. /«.■ Fishes of the Western
North Atlantic. H. B. Bigelow (ed). Mem. Sears Found. Mar. Res.
No. l(Pt. 3).
. 1963b. Family Albuhdae.p. 132-147. /n.Fishesofthe West-
em North Atlantic. H. B. Bigelow (ed). Mem. Sears Found. Mar.
Res. No. l(Pt. 3).
. 1963c. Family Engraulidae, p. 152-249. //;.■ Fishes of the
western North Atlantic. H. B. Bigelow (ed.). Mem. Sears Found.
Mar. Res. No. l(Pt. 3).
. 1963d. Family Clupeidae, p. 257-454. /«.■ Fishes of the west-
em North Atlantic. H. B. Bigelow (ed). Mem. Sears Found. Mar.
Res. No. KPt. 3).
, AND L. E. Cable. 1930. Development and life history of four-
teen teleostean fishes at Beaufort. N.C. Bull. Bur. Fish. Wash. 46
(l093):383-488.
, AND L. E. Cable. 1938. Further notes on the development
and life history of some teleosts at Beaufort, N.C. Bull. Bur. Fish.
Wash. 48(24):505-642.
-, AND W. C. ScHROEDER. 1928. Fishes of Chesapeake Bay. Bull.
Bur. Fish. Wash. 43(Pt. 1): 1-388.
Hipeau-Jacquotte. R. 1967. Notes de faunistique et biologic marine
de Madagascar. IV — Observations sur le comportement du poisson
Carapidae: Carapushomei (Kxchardson 1 844) de Madagascar. Reel.
Trav. Stn. Mar. Endoume Mars. Ease. Ser. Suppl. 6:141-151.
HoDA, S. M. S., AND H. Tsukahara. 1971. Studies on the develop-
ment and relative growth in the carp, Cypnnus carpio (Linne). J.
Fac. Agric. Kyushu Univ. 16:387-509.
Hoedeman, J. J. 1960a. Studies on callichthyid fishes. 3. Notes on
the development oi Callichthys ( 1 ) ( Pisces-Siluriformes). Bull. Aquat.
Biol. 1:53-72.
. 1960b. Studies on callichthyid fishes. 4. Development of the
skull of Callichthys and Hoplostenmm (1) (Pisces-Siluriformes).
Bull. Aquat. Biol. 1:73-84.
. 1960c. Studies on callichthyid fishes. 5. Development of the
skull of Callichthys and Hoplosternum (2) (Pisces-Siluriformes).
Bull. Aquat. Biol. 2:21-36.
. 1960d. Studies on callichthyid fishes. 6. The axial skeleton of
Callichthys and Hoplosternum (Pisces-Siluriformes). Bull. Aquat.
Biol. 2:37-44.
HoESE, D. F. 1971. A revision of the eastem Pacific species of the
gobiid fish genus Gohiosoma, with a discussion of relationships of
the genus. Unpubl. PhD dissertation. Univ. Calif, San Diego.
. 1976. A redescnption of HeterocUnus adelaide Costelnau
(Pisces: Clinidae) with description of related new species. Aust.
Zool. 19:51-67.
Hc>ESE, H. D., and R. H. Moore. 1977. Fishes of the Gulf of Mexico,
Texas, Louisiana, and adjacent waters. Texas A&M Univ. Press,
College Station and London.
HoFF, F. H.,Jr. 1970. Artificial spawning ofblack sea bass, Centropris-
tes slnatus melanus Ginsburg, aided by chorionic gonadotrophic
hormones. Fla. Dep. Nat. Resour. Mar. Res. Lab. Spec. Sci. Rep.
25.
HoLCiK, J. 1982. Towards the characteristics of the genera Hiicho and
Brachymystax {Pisces. Salmonidae). Folia Zool. 31:369-380.
HoLL, E. A. 1968. Notes of spawning behavior of barbel Clarias gar-
iepinus Burchell in Rhodesia. Zool. Afr. 3:185-188.
Hollister.G. 1936. Caudal skeleton of Bermuda shallow water fishes.
I. Order Isospondyli: Elopidae, Megalopidae, Albulidae, Clupeidae,
Dussumieriidae, Engraulidae. Zoologica (N.Y.) 21:257-290.
. 1937. Caudal skeleton of Bermuda shallow water fishes. II.
Order Percomorphi, suborder Percesoces: Atherinidae, Mugilidae.
Sphyraenidae. Zoologica (N.Y.) 22(3):265-279.
Holt, E. W. L. 1891. Survey of fishing grounds, west coast of Ireland,
1890. On the eggs and larvae of teleosteans. Sci. Trans. R. Dubl.
Soc, Ser. 2, 4:435-474.
, 1892. Rhombus maximus Linn. (The turbot). In: "Notes and
Memoranda." J. Mar. Biol. Assoc. U.K. 2:399-404.
. 1893. Survey of fishing grounds, west coast of Ireland, 1 890-
9 1 : on the eggs and post-larval stages of teleosteans. Sci. Trans. R.
Dubl. Soc, Ser. 2, 5.
. 1897. Preliminary notes on the reproduction of teleostean
fishes in the south-westem district. J. Mar. Biol. Assoc. U.K. 5:41-
50.
. 1898. Contributions to our knowledge of the plankton of the
Faeroe channel. No. V. Report on a collection of very young fishes
obtained by Dr. G. H. Fowler in the Faeroe Channel. Proc. Zool.
Soc. Lond. 1898:550-557.
HoNMA, Y. 1956. Short note on the breeding habit of a spiny lumpftsh.
Cyclolumpus aspemmus. Collect. Breed. 18:342-343.
. 1957. Notes on a smooth lumpsucker, Aplocyclus ventricosiis
(Pallas), during its breeding season. Collect. Breed. 19:235-236.
Horn, M. H. 1970a. The swimbladder as a juvenile organ in stro-
mateoid fishes. Breviora 359.
. 1970b. Systematics and biology of the stromateoid fishes of
the genus Pepnlus. Bull. Mus. Comp. Zool. Harv. Univ. 140:165-
262.
. 1973. Systematic companson of the stromateoid fishes Stro-
mateus brasiliensis Fowler and Slromateus sicllatus Cuvier from
coastal South America with a review of the genus. Bull. Br. Mus.
(Nat. Hist.) Zool. 24:317-339.
. 1975. Swim-bladder state and structure in relation to behavior
and mode of life in stromateoid fishes. Fish. Bull. U.S. 73:95-109.
, AND R. L. Haedrkh. 1973. Systematic and distributional
710
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
status of Psenes sio and Psenes pellucidus (Pisces, Stromateoidei)
in the eastern Pacific. Copeia 1973:167-169.
Hoss, D. E., AND J. H. S. Blaxter. 1982. Development and function
of the swimbladder-inner ear-lateral line system in the Atlantic
menhaden Brevoortia lyrannus {l^l^ohe). J. Fish Biol. 20: 1 3 1-142.
HoTTA, H. 1955. On the mature mugilid fish from Kabashima, Na-
gasaki Pref, Japan, with additional notes on the intestmal con-
volution of Mugilidae. Jpn. J. Ichthyol. 4:162-169.
, AND I.-S. Tung. 1972. Identification of fishes of the family
Mugilidae based on the pyloric caeca and the position on inserted
first intemeural spine. Jpn. J. Ichthyol., 14:62-66. (In Japanese.
English summary.)
HouDE. E. D. 1972a. Some recent advances and unsolved problems
in the culture of marine fish larvae. Proc. World Maricull. Soc. 3:
83-112.
. 1972b. Development and early life history of the northern
sennet, Sphyraena borealis DeKay (Pisces: Sphyraenidae) reared in
the laboratory. Fish. Bull. U.S. 70:185-195.
. 1974. Effects of temperature and delayed feeding on growth
and survival of larvae of three species of subtropical marine fishes.
Mar. Biol. (Bed.) 26:271-285.
. 1975. Effects of stocking density and food density on survival,
growth and yield of laboratory-reared larvae of sea bream Archo-
sargus rhomboidalts (L.) (Sparidae). J. Fish Biol. 7:1 15-127.
. 1977. Food concentrations and stocking density effects on
survival and growth of laboratory-reared larvae of bay anchovy
Anchoa mitchilti and lined sole Achirus lineatus. Mar. Biol. 43:
333-341.
. 1978. Critical food concentrations for larvae of three species
of subtropical marine fishes. Bull. Mar. Sci. 28:395-41 1.
. 1981. Distribution and abundance of four types of codlet
(Pisces: Bregmacerotidae) larvae from the eastern Gulf of Mexico.
Biol. Oceanogr. 1:81-104.
, AND J. D. Alpern-Lovdal. MS. Seasonality of occurrence,
foods and food preferences of ichthyoplankton in Biscayne Bay,
Florida.
, and p. L. Fore. 1973. Guide to identity of eggs and larvae
of some Gulf of Mexico clupeid fishes. Fla. Dep. Nat. Resour. Mar.
Res. Lab. Leafl. Ser. 4 (Pt. 1, No. 23).
— , C. R. FuTCH and R. Detwyler. 1970. Development of the
lined sole, Achirus lineatus. described from laboratory-reared and
Tampa Bay specimens. Fla. Dep. Nat. Resour. Mar. Res. Lab. Tech.
Ser. 62.
— , J. C. Leak, C. E. Dowd, S. A. Berkeley and W. J. Richards.
1979. Ichthyoplankton abundance and diversity in the eastern Gulf
of Mexico. Final Rep. to U.S. Bur. Land Mgt., Contract No. AA550-
CT7-28. Rosensteil School of Marine and Almosphenc Science.
Univ. Miami, FL.
— , and T. PoTTHOFF. 1976. Eggs and larval development of the
sea bream Archosargus rhomboidalts (Linnaeus): Pisces, Sparidae.
Bull. Mar. Sci. 26:506-529.
— , W. J. Richards AND V. P. Saksena. 1974. Description of eggs
and larvae of scaled sardine, Harengula jaguana. Fish. Bull. U.S.
72:1106-1122.
, AND R. C. ScHEKTER. 1978. Simulated food patches and Sur-
vival of larval bay anchovy Anchoa mitchilli and sea bream Ar-
chosargus rhomboidalis. Fish. Bull. U.S. 76:483-487.
— , AND L. J. SwANSON, Jr. 1975. Description of eggs and larvae
of yellowfin menhaden, Brevoortia smithi. Fish. Bull. U.S. 73:660-
674.
— , AND A. K. Taniguchi. 1979. Laboratory culture of marine
fish larvae and their role in marine environmental research, p. 176-
205. In: Advances in marine environmental research. Proceedings
of a Symposium. F. S. Jacoff(ed.). Environ. Res. Lab., U.S. Environ.
Prot. Agency, Narragansett, RI. Rep. No. EPA-600/9-79-035.
Howe, K. M., AND S. L. RICHARI5SON. 1978. Taxonomic review and
meristic variation in marine sculpins (Osteichthyes: Cottidae) of the
northeast Pacific Ocean. Fin. Rep.. NOAA-NMFS Contrib. No.
03-78-M02-120.
Howell, B. R. 1973. Marine fish culture in Britain VIII. A marine
rotifer Brachionus plicalilis Muller and the larvae of the mussel
Mylilus edults L. as food for larval flatfish. J. Cons. Int. Explor.
Mer 35:1-6.
HuBBS, C. 1952. A contribution to the classification of the blennioid
fishes of the family Clinidae. with a partial revision of the eastern
Pacific forms. Stanford Ichthyol. Bull. 4(2):41-I65.
. 1953a. Revision and systematic position of the blennioid
fishes of the genus NeocUnus. Copeia 1953:1 1-23.
. 1953b. Revision of the eastern Pacific fishes of the clinid genus
Labrisomus. Zoologica (N.Y.) 38:1 13-136.
. 1965. Developmental temperature tolerance and rates of four
southern California fishes, Fundulus parvipmnis. Athennops affinis.
Leureslhes tenuis, and Hvpsoblennius sp. Calif Fish Game 51:113-
122.
, andD. F. BuRNSiDE. 1972. Developmental sequences of Zrgo-
nectes notalus at several temperatures. Copeia 1972:862-865.
HuBBS, C. L. 1918. The fishes of the genus Atherinops. their variation,
distribution, relationships, and history. Bull. Amer. Mus. Nat. Hist.
38:409^40.
. 1925. The metamorphosis of the California ribbonfish Tra-
chipterus rex-salmonorum. Pap. Mich. Acad. Sci. Arts Lett. 5:469-
476.
. 1934. Coelorhynchus marinii, a new macrourid fish from Ar-
gentina and South Georgia. Occas. Pap. Mus. Zool. LIniv. Mich.
298:1-9.
. 1943. Terminology of early stages of fishes. Copeia 1943:260.
. 1945. Phylogenetic position of the Cithandae. a family of
flatfishes. Univ. Mich. Mus. Zool. Misc. Publ. 63.
. 1946. Pertinence of the East Indian heterosomate fish genus
Lepidoblepharon to the Citharidae. Copeia 1946:97.
. 1951. Identification as osmerids of two fishes described from
the North Pacific as paralepidids. Copeia 1951:296.
. 1953. Synonymy of the bathypelagic fish genus Rhynchohy-
alus. referred to the expanded family Argentinidae. Copeia 1953:
96-97,
. 1958. Dikellorhynchus and Kanazawaichlhys: nominal fish
genera interpreted as based on prejuveniles of Malacanthus and
Antennarius, respectively. Copeia 1958:282-285.
, AND Y. T. Chc. 1934. Asiatic fishes (Diplopnon and Laeops)
having a greatly elongated dorsal ray in very large post-larvae.
Occas. Pap. Mus. Zool. Univ. Mich. 299:1-7.
, AND L. C. HuBBS. 1945. Bilateral asymmetry and bilateral
variation in fishes. Pap. Mich. Acad. Sci. Arts Lett. 30:229-310.
, AND T. IwAMOTO. 1977. A new genus {Mesobnis). and three
new bathypelagic species of Macrouridae (Pisces, Gadiformes) from
the Pacific Ocean. Proc. Calif Acad. Sci. Fourth Ser. 41(7):233-
251.
, AND E. M. Kampa. 1946. The early stages (egg, prolarva and
juvenile) and the classification of the California flying fish. Copeia
1946:188-218.
, AND R. L. WiSNER. 1964. fam/w.v. a new genus of myctophid
fishes from the northeastern Pacific, with two new species. Zool.
Meded. (Leiden) 39:445-463.
— , AND . 1980. Revision of the sauries (Pisces, Scomber-
esocidae) with descriptions of two new genera and one new species.
Fish Bull. U.S. 77:521-566.
Hudson, L. L., and J. D. Hardy, Jr. 1975. Eggs and larvae of the
Atlantic seahorse. Hippocampus hudsonnis. Univ. Md., CEES, Ref
175-I2CBL.
Hlighes, D. R. 1981. Developmeni and organization of the posterior
field of ctenoid scales in the Platycephalidae. Copeia 1981:596-
606.
HuLET, W. H. 1978. Structure and functional development of the eel
leptocephalus Ariosoma balearicum (De La Roche, 1809). Philos.
Trans. R. Soc. Lond. 282(987): 107-1 38.
Hulley, p. a. 1972. A report on the mesopelagic fishes collected
LITERATURE CITED
711
during the deep-sea cruises on R.S. "Africana II," 1961-1966. Ann.
S. Afr. Mus. 60:197-236.
. 1981. Results of the research cruise of FRV "Walter Herwig"
to South America. LVIII. Family Myctophidae(Osteichthyes, Myc-
tophiformes). Arch. Fischereiwiss. 3 1(1): 1-300.
, AND R. E. Rau. 1969. A female Regalecus glesne from Cape
Province, South Africa. Copeia 1969:835-839.
HuMASON, G. L. 1979. Animal tissue techniques. W. H. Freeman and
Co., San Francisco.
Humphries, J. M., F. L. Bookstein, B. Chernoff, C. R. Smith, R. L.
Elder AND S. G. Poss. 1981. Multivariate discrimination by shape
in relation to size. Syst. Zool. 30:291-308.
Hunter, C. J. 1969. Confirmation of symbiotic relationship between
liparid fishes and male king crab. Pac. Sci. 23:546-547.
Hunter, J. R. 1972. Swimming and feeding behavior of larval an-
chovy Engraulis mordax. Fish. Bull. U.S. 70:821-838.
. 1976. Culture and growth of northern anchovy Engraulis
mordax \aT\?ie. Fish. Bull. U.S. 78:81-88.
. 1981. Feeding ecology and predation of marine fish larvae, p.
a-ll. In: Marine fish larvae. Morphology, ecology and relation
to fisheries. R. Lasker (ed.). Univ. Washington Press, Seattle. WA.
. 1984. Inferences regarding predation on the early life stages
of cod and other fishes, p. 533-562. In: The propagation of cod.
E. Dahl, D. S. Danielssen, E. Moksness and P. Solemdal (eds.).
Flodevigen Rapp. 1.
, AND K. M. Coyne, 1982. The onset of schooling in northern
anchovy larvae Engraulis mordax. Calif. Coop. Oceanic Fish. In-
vest. Rep. 23:246-251.
, S. E. Kaupp AND J. T. Taylor. 1982. Assessment of effects
of UV radiation on marine fish larvae, p, 459-497, In: The role of
solar ultraviolet radiation in marine ecosystems. NATO Conference
Series IV: Marine Sciences. J. Calkins (ed.). Plenum Press, New
York.
AND C. A. Kimbrell. 1980. Egg cannibalism in the northern
anchovy Engraulis mordax. Fish. Bull. U.S. 78:81 1-816.
HuREAu, J. C. 1971. Notes sur la famille des Congiopodidae (Te-
leosteens, Perciformes): redecouverte de Zanctorhynchus spinifer
(Gunther, 1880) aux iles Kerguelen et rehabilitation de Congio-
podus kienernSauvage, 1878). Bull. Mus. Natl. Hist. Nat. 42: 1019-
1026.
, AND A. ToMO. 1977. Bovichthys elongalus n. sp., Poisson
Bovichthyidae, famille nouveau pour I'Antarctique, Cybium 3e Ser.
1977(l):67-74.
HussAiN, N,, S, AiovTSU AND C. El-Zahr. 1981. Spawning, egg and
early development, and growth oi Acanlhopagrus cmieri (Sparidae).
Aquaculture 22:125-136.
HussAiN, N. A., AND M. HioucHi, 1980. Larval rearing and devel-
opment of the brown spotted grouper, E^inephelus tauvina (For-
skal). Aquaculture 19:339-350.
Huxley, J. S. 1932. Problems of relative growth. Methucn and Co.,
London.
HuYOEBAERT, B., AND D. Nolf. 1979. An annotated bibliography of
paleontological and systematic papers on fish-otoliths, published
since 1968. Meded. Werkgr. Tert. Kwart. Geol. 16(4): 139-170.
Ida, H. 1976. Removal of the family Hypoptychidae from the sub-
order Ammodytoidei, order Perciformes, to the suborder Gaster-
osteoidei, order Syngnathiformes. Jpn. J. Ichthyol. 23:33-42.
Iljin, B. S. 1930. Le systeme des Gobiides. Trab. Inst. Esp. Oceanogr.
(2): 1-63.
Imai, S. 1 94 1 . Seven new deep-sea fishes obtained in Sagami Bay. Jpn.
J. Zool. 9(2):233-250.
. 1958. Young of Myctophum orienlale (Gilbert), p. 21. In:
Studies on the eggs, larvae and juveniles of Japanese fishes. Ser. I.
Uchida el al. (eds.). Fac. .Agric. Kyushu Univ., Fukuoka, Japan.
(In Japanese.)
. 1959. Studies on the life histories of the flying-fishes found in
theadjacent waters of Japan. — I. Mem. Fac. Fish. KagoshimaUniv.
7:1-85. (In Japanese.)
. 1960. Studies on the life histories of flying-fishes found in the
adjacent waters of Japan — II. Mem. Fac. Fish. ICagoshima Univ.
8:8-45. (In Japanese.)
, AND K. Nakahar.a. 1957. Coreoperca kawamurainnd'WsWk
history, p. 591-601. In: Syehiro et al. (eds.). Suisangaku Shusei.
(In Japanese.)
Inaba, D.,M. Nomura, AND M. Nakamura. 1957. Preliminary report
on the spawning of grass-carp and silver-carp in the Tone River,
Japan and the development of their eggs. J. Tokyo Univ. Fish.
43(l):81-96.
Inada, T. 1981a. Two nominal species of A/er/«cc/!tj from New Zea-
land and southern South America. Jpn. J. Ichthyol. 28:31-36.
. 1981b. Studies on the merlucciid fishes. Bull. Far Seas Fish.
Res. Ub. (Shimizu) 18:1-172.
, and I. Nakamura. 1975. A comparative study of two pop-
ulations of the gadoid fish Micromesislius auslralis from the New
Zealand and Patagonian-Falkland regions. Bull. Far. Seas Fish. Res.
Lab. (Shimizu) 13:1-26.
Inglis, W. G. 1966. The observational basis of homology. Syst. Zool.
15:219-228.
Innes, W. T. 1956. Exotic aquarium fishes. 19th Ed. G. S. Myers (ed.).
Innes Publ. Co., Phila.
Irie, T. 1960. The growth of the fish otolith. J. Fac. Fish. .'\nim. Husb.
Hiroshima Univ. 3:203-221.
Ishigaki, T. 1957. A few informations concerning "Shiwa-ikanago,
Hypoptychus dybowskii Steindachner", in the waters around Hok-
kaido. Month. J. Hokkaido Fish Res. Stn. 14(8): 14-24. (In Jap-
anese.)
Ishiyama, R, 1951. Revision of the Japanese mugilid fishes especially
based upon the osteological characters of the cranium. Jpn. J. Ich-
thyol. 1:230-250. (In Japanese. English summary.)
Ivankov, V. N., AND V. P. Kurdiayeva. 1973. Systematic differences
and the ecological importance of the membranes in fish eggs. J.
Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 13:864-873.
IvANTSOFF, W. 1978. Taxonomic and systematic review of the Aus-
tralian fish species of the family Atherinidae with references to
related species of the Old World. Unpubl. PhD dissertation. Mac-
quarie University, Sydney.
Iwami, T,, andT. Abe. 1980. Records of adults of some scopelarchid
fishes from the western North Pacific and the Southern Ocean, with
osteological notes on five species of the genus Benthalbella. U. O.
No. 31:1-20.
IwAMOTO, T. 1966. The macrourid fish Cetonurus glohiceps in the
Gulf of Mexico. Copeia 1966:439-442.
. 1974, Nezumia {Kuronezuniia) bubonis. a new subgenus and
species of grenadier (Macrouridae: Pisces) from Hawaii and the
western North Atlantic. Proc. Calif Acad. Sci., Fourth Ser.. 39(22):
507-516.
. 1978. Eastern Pacific macrourids of the genus Coelorhynchus
Gioma (Pisces: Gadiformes), with description of a new species from
Chile. Proc, Calif Acad. Sci., Fourth Ser., 41(12):307-337.
. 1979. Eastern Pacific macrourine grenadiers with seven bran-
chiostegal rays (Pisces: Macrouridae). Proc. Calif Acad. Sci. 42(5):
135-179.
, AND J. C. Staiger. 1976. Percophidid fishes of the genus
Chrionema Gilbert. Bull. Mar. Sci. 26:488-498.
, AND D. L. Stein. 1974. A systematic review of the rattail fishes
(Macrouridae: Gadiformes) from Oregon and adjacent waters. Oc-
cas. Pap. Calif Acad. Sci., No, 1 1 1:1-79.
Jacino, A. 1909. Uovo e larve di rraf/jvp^erwisp. Arch. Zool. Napoli
3(4):479-484.
Jackson, P. D. 1978. Spawning and early development of the river
blackfish, Gadopsis marmoralus Richardson (Gadopsiformes: Ga-
dopsidae) in the McKenzie River, Victoria. Aust. J. Mar. Fresh-
water Res. 29:293-298.
James, G.D. 1976a, Eggs and larvae o( Ihe IrevaWy Caran.xgeorgianus
(Teleostei: Carangidae). N.Z. Mar. Freshwater Res. 10:301-310.
. 1976b. Cyllus iraversi Hullon: Suveni\e (onn of C. venlralis
Barnard and Davies (Pisces, Zeidae). J. R. See. N.Z. 6(4):493-498.
712
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
James, P. S. B. R. 1970. Micrognathus brevirosrns (Ruppell) (Family
Syngnathidae: Pisces)— a new record from the Indian Seas with
observations on its early development. J. Mar. Biol. Assoc. India
12:158-162.
Janssen, J., AND G. R. Harbison. 1981. Fish in salps: the association
of squaretails (Tetragonurus spp.) with pelagic tunicates. J. Mar.
Biol. Ass. U. K. 61:917-927.
Jenkins, G. P., N. E. Milward and R. F. Hartwick. 1984. Identi-
fication and description of larvae of Spanish mackerels, genus
Scomberomorus (Teleostei: Scombridae), in shelf waters of the Great
Bamer Reef. Aust. J. Mar. Freshwater Res. 35:341-353.
Jensen, A. S. 1935. The Greenland halibut (Reinhardtius hippoglos-
soides). its development and migrations. K. Dan. Vidensk. Selsk.
Skr. 9(4): 1-32.
. 1 948. Contributions to the ichthyofauna of Greenland. Spolia
Zool. Mus. Haun. 9:1-176.
Jensen, R.J. 1981. Wagner networks and Wagner trees: a presentation
of methods for estimating parsimonious solutions. Taxon 30:576-
590.
Jerzmansica, A. 1968. Ichtyofaune des couches a menilite (flysch des
Karpathes). Acta Palaeontol. Pol. 13(3):379-488.
Jespersen, p. 1942. Indo-Pacific leplocephalids of the genus Angiulta.
Systematic and biological studies. Dana-Rep. Carlsberg Found. 22.
, AND A. V. TAning. 1919. Some Mediterranean and Atlantic
Stemoptychidae. Preliminary Note. Saertryk af Vidensk. Medd.
Dansk Naturh. Foren. 70:215-226.
, AND . 1926. Mediterranean Stemoptychidae. Rep. Dan.
Oceanogr. Exped. Mediterr. 2 (A. 12): 1-59.
Jhingran, V. G. 1975. Fish and fisheries of India. Hindustan Pub-
lishing Corporation. Delhi, India.
Jillett, J. B. 1 968. The biology of Acanlhoclinus quadndactylus (Bloch
and Schneider) (Teleostei — Blennioidea). II. Breeding and devel-
opment. Aust. J. Mar. Freshwater Res. 19:9-18.
Job, T. J. 1940. On the breeding and development of Indian "mos-
quito-fish" of the genera Aplocheilus McCelland and Oryzias Jordan
and Snyder. Rec. Indian Mus. (Calcutta) 42:51-79.
, AND S. Jones. 1938. Studies on the development of the Indian
garfish, Tylosurus sirongylunis (van Hass.) with notes on the young
stages of Hemirhamphus gaimardi Cuv. & Val. Rec. Indian Mus.
(Calcutta) 40(3):245-253.
Johansen, F. 1912. The fishes of the Danmark Expedition. Medd.
Gronl. 45(12):633-675,
John, C. M. 1950. Early stages in the development of the sucker fish
Echeneis naucrales Linn. Bull. Cent. Res. Inst. Univ. Travancore
India lC(l):47-55.
John, M. A. 1951a. Pelagic fish eggs and larvae of the Madras coast.
J. Zool. Soc. India 2:267-268.
. 1951b. Pelagic fish eggs and larvae of the Madras coast. J.
Zool. Soc. India 3:41-69.
Johnsen, S. 1921. Ichthyologiske notiser I. Bergens Mus. Aarbok 1918-
1919. Naturvid. Raekke No. 6. (English summary.)
. 1927. On some balhypelagic stages of macrourid fishes. Nytt
Mag. Natur\'id. 65:221-241.
Johnson, C. R. 1 969. Contnbutions to the biology of the showy snail-
fish, Liparis pulchellus (Lipahdae). Copeia 1969:830-835.
, D. A. Ratkowskv AND R. W. G. White. 1983. Multivariate
analysis of the phenolypic relationships of the species of Paraga-
laxias and Galaxias (Pisces: Galaxiidae) in Tasmania. J. Fish Biol.
23:49-63.
, R. W. G. White AND Y. A. E. Beck. 1981. Cytotaxonomy of
seven species of Galaxias (Pisces: Galaxiidae) in Tasmania. Ge-
netica 56:17-21.
Johnson, G. D. 1975. The procurrent spur, an undescribed perciform
character and its phylogenetic implications. Occas. Pap. Calif. Acad.
Sci. 121:1-23.
. 1978. Development of fishes of the Mid-Atlantic Bight. An
Atlas of egg. larval and juvenile stages. Vol. IV. Carangidae through
Ephippidae. U.S. Fish Wildl. Serv. Biol. Serv. Program FWS/OBS-
78/12.
. 1980. The limits and relationships of the Lutjanidae and as-
sociated families. Bull. Scripps Inst. Oceanogr. Univ. Calif 24.
. 1983. Niphon spinosus: & primitive epinephelineserranid, with
comments on the monophyly and intrarelationships of the Serran-
idae. Copeia 1983:777-787.
, AND P. Keener. 1984. Aid to identification of Amencan grou-
per larvae. Bull. Mar. Sci. 34:106-134.
Johnson, R. K.. 1970. A second record of Korsogasler nanus Parr
(Beryciformes: Korsogasteridae). Copeia 1970:758-760.
. 1974a. A Macristium larva from the Gulf of Mexico with
additional evidence for the synonymy of Macnslntm with Balhy-
iflMrwi (Myctophiformes: Bathysauridae). Copeia 1974:973-977.
. 1974b. A revision of the alepisauroid family Scopelarchidae
(Pisces: Myctophiformes). Fieldiana Zool. 66.
. 1975. A new myctophid fish. Bolinichthys distofax. from the
western and central north Pacific Ocean, with notes on other species
o( Bolinichthys. Copeia 1975:53-60.
. 1982. Fishes of the families Evermannellidae and Scopelar-
chidae: systematics, morphology, interrelationships and zoogeog-
raphy. Fieldiana Zool. N.S. 12.
— , and D. M. Cohen. 1974. Results of the research cruises of
FRV "Walther Herwig" to South America. 30. Revision of the
chiasmodontid fish genera Dysalotus and Kali, with descriptions
of two new species. Arch. Fischereiwiss. 25:13-46.
— , and R. M. Feltes. 1984. A new species of r/wc/^erna (Sal-
moniformes: Photichthyidae) from the Red Sea and Gulf of Aqaba,
with comments on the depauperacy of the Red Sea mesopelagic
fish fauna. Fieldiana Zool. N. S. 22.
— , AND D. W. Greenfield. 1976. A new chaenopsid fish. Em-
blemana hylloni. from Isla Roatan, Honduras. Fieldiana Zool. 70(2):
13-28.
Johnson, S. E. 1918. Osteology of the gruntfish, Rhamphocotlus nch-
ardsoni. J. Morphol. 31:461-477.
JoLicoEUR. P. 1963. The multivanate generalization of the allometry
equation. Biometrics 19:497-499.
Jolly, V. H. 1967. Observations on the smelt Relropinna lacustris
Stokell. N.Z. J. Sci. 10:330-355.
Jones, A. 1972. Studies on egg development and larval rearing of
turbot, Scophthalmus maximus L.. and bnll Scophthalmus rhom-
bus L., in the laboratory. J. Mar. Biol. Assoc. U.K. 52:965-986.
Jones, A. C. 1962. The biology of the euryhaline fish Leptocollus
armatus armalus Girard. Univ. Calif Publ. Zool. 67(4):32 1-368.
Jones, G. G., and M. A. Tabery. 1980. Larval development of the
banded killifish (Fundulus diaphanus) with notes on the distribu-
tion in the Hudson River estuary, p. 25-35. In: Proc. Fourth Ann.
Urval Fish Conf L. A. Fuiman (ed.). U.S. Fish Wildl. Serv., Na-
tional Power Plant Team, Ann Arbor, Mich., FWS/OBS-80/43.
Jones, M. P.. F. G. T. Holliday and A. E. G. Dlinn. 1966. The
ultrastructure of the epidermis of larvae of the herring (Clupea
harengus) in relation to the rearing salinity. J. Mar. Biol. Assoc.
U.K. 46:235-239.
Jones, P. W., F. D. Martin and J. D. Hardy, Jr. 1978. Development
of fishes of the mid-Atlantic bight, an atlas of egg. larval, and
juvenile stages. Volume 1. Acipenseridae through Ictaluridae. U.S.
Fish Wildl. Serv., Biol. Serv. Prog. FWS/OBS-78/12.
Jones, R., and W. B. Hall. 1974. Some observations on the popu-
lation dynamics of the larval stage in the common gadoids, p. 87-
102. /«.■ The early life history offish. J. H. S. Blaxter (ed.). Springer
Verlag, Berlin.
Jones, S. 1937. Observations on the breeding habits and development
of certain brackish water fishes of Adyar, Madras. Proc. Indian
Acad. Sci. Sect. B 5:261-289.
. 1938. On the breeding habits and development of a cypnnid,
Danio (Danio) malabaricus (Jerdon) in Ceylon. Ceylon J. Sci. Sec.
C 6:79-90.
LITERATURE CITED
713
, AND M. Kljmaran. 1962. Notes on eggs, larvae and juveniles
of fishes from Indian waters. Xll. Myriprtslis inurdjan. Xlll. Hol-
ocenlrus %p. Indian J. Fish. Sect. A 9( 1): 155-167.
.AND . 1964. On the fishes of the genus .S'c7;»!(y/£'rM Gil-
tay from the Indian Ocean. J. Mar. Biol. Assoc. India 6:257-264.
, AND . 1967. Notes on eggs, larvae, and juveniles of
fishes from Indian waters. XV. Pegasus volilans Linnaeus. XVI.
Daaytoptena onentalis (Cuvier and Valenciennes). XVII. Dacty-
loptena macracanlhus Bleeker). Indian J. Fish. 1 1:232-246.
.AND . 1950. Spawning habits and development of the
Gangetic anchovy, Selipinna phasa (Hamilton). Curr. Sci. 19:25.
, AND . 1951. Notes on the bionomics and developmental
stages of some Indian flatfishes. J. Zool. Soc. India 3:71-83.
, AND . 1952. Observations on the development and
systematics of the fishes of the genus Coilia Gray. J. Zool. Soc.
India 4: 17-36.
. AND V. B. Pantulu. 1958. On some larval and juvenile fishes
from the Bengal and Orissa coasts. Indian J. Fish. 5:1 18-143.
Jordan. D. S. 1892. Relations of temperature to vertebrae among
fishes. Proc. U.S. Natl. Mus. 14:107-120.
. 1923. A classification of fishes including families and genera
as far as known. Stanford Univ. Publ. Biol. Sci. 3(2):77-243.
. and B. W. Evermann. 1896-1900. The fishes of North and
Middle Amenca. U.S. Natl. Mus. Bull. 47.
, AND C. H. Gilbert. 1899. The fishes of the Bering Sea. p.
433-492. In: Fur-seals and fur-seal islands of the North Pacific
Ocean, Pt. 3. D. S. Jordon (ed.). Washington, D.C.
, AND E. C. Starks. 1904. A review of the Cottidae or sculpins
found in the waters of Japan. Proc. U.S. Natl. Mus. 27:231-335.
JiDE. D. J. 1982a. Family Cypnnodontidae, p. 477-486. In: Identi-
fication of larval fishes of the Great Lakes basin with emphasis on
the Lake Michigan drainage. N. A. Auer (ed.). Great Lakes Fish.
Comm. Spec. Publ. 82-3.
. 1982b. Family Gadidae, codfishes, p. 470-476. In: Identifi-
cation of larval fishes of the Great Lakes basin with emphasis on
the Lake Michigan drainage. N. A. Auer (ed.). Great Lakes Fish.
Comm. Spec. Publ. 82-3.
JiiNGERSEN.H. F. E. 1910. Ichthyotomical contributions. II. The struc-
ture of the Aulostomidae, Syngnathidae and Solenostomidae. K.
Dan. Vidensk. Selsk, Skr. Ser. 8, 7:268-364.
K.ADAM, K. M. 1961. The development of the skull in /Verop/j/i. Acta.
Zool. (Stockh.) 42:257-298.
Kahan, D., G. Uhlig, D. Schwenzer and L. Horowitz. 1982. A
simple method for cultivating harpacticoid copepods and offering
them to fish larvae. Aquaculture 26:303-310.
K,MLL, W. M. 1967. Ecology and behavior of the cyprinodontid fishes
Jordanellaflohdae Goode and Bean. Floridichthys carpio (Gunther)
and Cypnnodon variegatus Lacepede. LIrhpubl. PhD dissertation.
Cornell Univ., Ithaca, NY.
Kaiser, C. E. 1973. Age and growth of horse mackerel Trachiinis
mMrp/iy/ off" Chile. Trans. Am. Fish. Soc. 102(3):591-595.
Kamlva, T. 1925. The pelagic eggs and larvae of fishes in the Tateya-
ma Bay (Pref Chiba) 3rd Report. J. Imp. Fish. Inst. (Tokyo) 21(3):
27-35.
Kanazawa, R. H. 1952. More new species and new records of fishes
from Bermuda. Fieldiana Zool. 34(7):7I-100.
Kanidev, A. N. 1961. Embryonic development of .4//u'n«a /(epienis
and A. mochon pontica. Tr. Karadag. Biol. Stn. 17:24-45.
Karamchandani, S. J., AND M. P. MoTWANi. 1955. Early life-history,
bionomics and breeding of /?«a r;7a (Hamilton). J. Zool. Soc. India
7:115-126.
.and . 1956. On the larval development of four species
of freshwater cat fishes from the River Ganga. J. Zool. Soc. India
8(1): 19-34.
Karrer, C. 1976. Uber drei mesopelagische fisharten aus dem Golf
von Guinea. Mitt. Zool. Mus. Berl. 52:177-182.
. 1982. Anguilliformes du Canal de Mozambique (Pisces. Te-
leostei). Faune Trop. 23:1-1 16.
Katayama, M. 1960. Fauna Japonica Serranidae. (Pisces). Tokyo
News Serv., Ltd.. Tokyo.
Kaufman, L. S.. and K. F. Liem. 1982. Fishes of the suborder La-
broidei (Pisces: Perciformes): phylogeny, ecology, and evolutionary
significance. Breviora (472): 1-1 9.
K^WAGLiCHi, K., AND K. Aioi. 1972. Myctophid fishes of the genus
Myctophum (Myctophidae) in the Pacific and Indian Oceans. J.
Oceanogr. Soc. Jpn. 28:161-175.
, AND J. L. Butler. In press. Fishes of the genus Nansenia
(Microstomatidae) with descriptions of seven new species. Contrib.
Sci. (Los Ang.).
, H. Ikeda, M. Tamura and S. Ueyanagi. 1972. Geographical
distribution of surface migrating myctophid fishes (genus Mycto-
phum) in the tropical and subtropical Pacific and Indian Oceans.
Bull. Far. Seas Fish. Res. Ub. (Shimizu) 6:23-37.
, and R. M ARUMO. 1967. Biology of Gonostoma gracile (Gon-
ostomatidae). 1. Morphology, life history and sex reversal, p. 253-
269. In: Information Bulletin on Planktology in Japan. Commem-
oration Number of Dr. Y. Matsue.
. AND J. Mauchline. 1982. Biology of myctophid fishes (family
Myctophidae) in the Rockall Trough. Northeastern Atlantic Ocean.
Biol. Oceanogr. 1:337-373.
Ka-i A, C. M., A. E. DizoN, S. D. Hendrix. T. K. Kaza.ma and M. K.
K. Queenth. 1982. Rapid and spontaneous maturation, ovula-
tion, and spawning of ova by newly captured skipjack tuna, Kal-
suwonus pelamis. Fish. Bull. U.S. 80:393-396.
Keene, M. 1973. Mclamphaes ehelingi. a new species of beryciform
fish (Melamphaidae) from the North Atlantic. Copeia 1973:794-
800.
Kelvin, T. M., G. Phillips, J. E. Thomerson and D. C. Taphorn.
1980. Egg chorion ornamentation and taxonomy of the annual
killifishes Auslrofundutus. Rachovia. and Terranalos. Am. Soc. Ich-
thyol. Herpetol., 60lh Annual Meeting. Fort Worth. Texas. (Ab-
stract.)
Kelley, S. MS. A description of laboratory-reared larvae of the labrid
fish, Lachnolaimus maximus.
Kendall, A. W., Jr. 1972. Descnption of black sea bass, Cen/ropmnj
striata (Linnaeus), larvae and their occurrence north of Cape Look-
out. North Carolina, in 1966. Fish. Bull. U.S. 70:1243-1260.
. 1976. Predorsal and associated bones in serranid and gram-
mistid fishes. Bull. Mar. Sci. 26:585-592.
. 1977. Relationships among American serranid fishes based
on the morphology of their larvae. Unpubl. PhD dissertation. Univ.
Calif San Diego, San Diego, CA.
. 1979. Morphological comparisons of North American sea bass
larvae (Pisces: Serranidae). NOAA Tech. Rep. NMFS Circ. 428.
. 1982. Elucidating relationships among fishes using larval char-
acters. Mimeogr.
, AND M. P. FaHAY.
1979. Larva of the serranid fish Gonio-
plectrus hispanus with comments on its relationships. Bull. Mar.
Sci. 29:117-121.
, AND D. Gordon. 1981. Growth rate of Atlantic mackerel
(Scomber scombrus) larvae in the middle Atlantic Bight. Rapp. P.
V. Reun. Cons. Int. Explor. Mer 178:337-341.
, C. D. Jennings, T. M. Beasley, R. Carpenter and B. L. J.
SoMAYAjuLU. 1983. Discovery of a fish (zoarcid) egg cluster bur-
ied 10-12 cm deep in Washington continental slope sediments.
Mar. Biol. 74:193-200.
, andB. ViNTER. 1984. Development of hexagrammids (Pisces:
Scorpaeniformes) in the Northeastern Pacific Ocean. NOAA Tech.
Rep. NMFS 2.
Kendall, W. C, and E. L. Gole^sborough. 1911. Reports on the
scientific results of the expedition to the tropical Pacific, in charge
of Alexander Agassiz, by the U.S. Fish Commission Steamer "Al-
batross," from August 1899 to March 1900. 13. The Shore fishes.
Mem. Mus. Comp. Zool. Harvard 26(7):293-344.
Kennedy, M., and T. Champ. 1971. Occurrence of eggs of Echiodon
714
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
drummondi Thompson on the coast of County Kerry. Ir. Fish.
Invest. Ser. B (Mar.)7:55-59.
AND p. FiTZMAURicE. 1969. Pelagic eggs and young stages of
fishes taken on the south coast of Ireland in 1967. Ir. Fish. Invest.
Ser. B (Mar.)5:5-36.
Khaldinova, N. a. 1936. Study of spawning, eggs, and larvae of
Elegmus (Gadus) namga Pall. Zool. Zh. 15(2):32 1-339. (In Rus-
sian, English summary.)
Khan, N. Y. 1972. Comparative morphology and ecology of the pe-
lagic larvae of nine Cottidae (Pisces) of the northwest Atlantic and
St. Lawrence drainage. Unpubl. PhD thesis. Univ. Ottawa, Canada.
AND D. J. Faber. 1974. A comparison of larvae of the
deepwater and fourhom sculpin. Myoxocephalus quadricorms L.
from North Amenca. 1 . Morphological development, pp. 703-7 1 2.
In: The early life history offish. Blaxter, J. H. S. (ed). Springer-
Verlag, New York.
KiDO, K. 1983. New and rare liparidid species from the Okhotsk and
Bering Seas and their adjacent waters. Jpn. J. Ichthyol. 29:374-
384.
Kjliachenkova, V. A. 1970. Development and distribution of eggs
and larvae of Trachurus Irachurus L. Rapp. P. V. Reun. Cons. Int.
Explor. Mer. 159:194-198.
Kjm, Y. U., P. Chin, T. Y. Lee and Y. J. Kang. 1981. Studies on the
fish larvae of coastal waters in Korea. Publ. Inst. Mar. Sci. Natl.
Fish. Univ. Busan 13:1-35.
Kindred, J. E. 1921. The chondrocranium of Synanathus fuscus. J.
Morphol. 35:425-456.
KiNG, D. P. F., M, J. O'Toole and A. A. Robertson. 1977. Early
development of the south African maasbanker, Trachurus trachu-
rus, at controlled temperatures. Fish. Bull. S. Afr. 9:16-22.
Kjnne, O. 1977. Research Cultivation: (13) Pisces, p. 968-1035. In:
Marine ecology. Vol. Ill, part 2. O. Kinne (ed.). John Wiley & Sons,
Chichester.
Kirschbaum, F. 1984. Reproduction of weakly electric teleosts: just
another example of convergent development? Environ. Biol. Fishes
10:3-14.
, and J. p. Denizot. 1975. Sur la differentiation des electro-
recepteurs chez Marcusenius sp. (Mormyrides) et Eigenmanma vi-
rescens (Gymnotoides), poissons electriques a faible decharge. C.
R. Hebd. Sciences Acad. Sci. Ser. D. Sci. Nat 28:419-422.
, and G. W. M. Westby. 1975. Development of the electric
discharge in mormyrid and gymnotid fish {Marcusenius sp. and
Eigenmannia virescens). Experientia (Basel) 31:1290-1293.
Klawe, W. L. 1960. Animals from night-light collections. In: Physical,
chemical, and biological observations in the eastern Tropical Pacific
Ocean — Scot Expedition, April-June 1958. R. W. Holmes and M.
Blackburn (eds.). U.S. Fish Wildl. Serv. SSRF 345.
Klee, a. J. 1982. Comments on the genus Rivulus. J. Am. Killifish
Assoc. 15(3):76-79.
Klein-Macphee, G., W. H. Howell and A. D. Beck. 1982. Com-
parison of a reference strain and four geographical strains of Artemia
as food for winter flounder (Pseudopleuronectes americanus) larvae.
Aquaculture 29:279-288.
Kliever, R. G. 1976. Natural history of May nea californica (Pisces:
Zoarcidae) in a drift seaweed habitat in the Monterey Submarine
Canyon, Monterey Bay, California. Unpubl. MA Thesis. San Jose
State Univ., San Jose, CA.
Kluge, A. G., AND J. S. Farris. 1969. Quantitative phyletics and the
evolution of anurans. Syst. Zool. 18:1-32.
Kobayashi, K. 1957. Larvae and young of the sable fish, ,-l«op/opoOTa
fimbria (Pallas), from the sea near the Aleutian Islands. Bull. Jpn.
Soc. Sci. Fish 23:376-382. (In Japanese. English abstract.)
. 1961a. Young of the wolflSsh, Anarchichas orientalis (Pallas).
Bull. Fac. Fish. Hokkaido Univ. 12(l):l-4. (In Japanese, English
abstr.)
. 1961b. Larvae and young of the quillfish, Plilichlhys goodei
(Bean), from the Okhotsk Sea. Bull. Fac. Fish. Hokkaido Univ.
12(l):5-8. (In Japanese, English abstr.)
-. 1961c. Larvae and young of the sand lance, .-I »imo(/i7e.s /le.v-
aplerus (Pallas) from the nonh Pacific. Bull. Fac. Fish. Hokkaido
Univ. 12:111-120.
. 1 962. Lai-vae of the smooth lumpsucker, Aptocychis venlri-
cosus (Pallas), with discussion on revision of the taxonomy of the
species. Bull. Fac. Fish. Hokkaido Univ. 13(3): 153-1 64.
, AND K. Abe. 1962. Studies on the larvae and young of fishes
from the boundary zones off' the south-eastern coast of Hokkaido,
Japan. Bull. Fac. Fish. Hokkaido Univ. 13:165-179. (In Japanese,
English summary.)
, andM. Igarashi. 1961. Larvae and young of GV /-('//a pKMtVa/a
Gray from Hakodate Bay in southern Hokkaido. Bull. Fac. Fish.
Hokkaido 12:121-127. (In Japanese, English abstract.)
, M. Mikawa, AND J. Ito. 1968. Descriptions of the young and
one, immature adult specimen of the coster dory, Altocyllus ver-
rucosus (Gilchrist) from the northern pari of the Pacific. Bull. Fac.
Fish. Hokkaido Univ. 19(1): 1-5.
Kobayashi, T., Y, Dotsli and N. Miura. 1973. Egg development
and rearing experiments of the larvae of the mud skipper, Per-
iophthalmus cantonensis. Bull. Fac. Fish. Nagasaki Univ. 33:49-
62.
Kobayasi, H., and S. Moriyana. 1957. On the development of a
loach, Barbalula toni areas (Jordan et Fowler). Jpn. J, Ichthyol.
6:177-183.
Kobelkowsky, A. D., and A. M. Resendez. 1972. Estudio compar-
ativo del endosqueleto de Mugil cephahis y Mugil curema (Pisces,
Perciformes). An. Inst. Biol. Univ. Nac. Auton. Mex. Ser. Cienc.
Mar. Limnol. 43(1):33-81.
KoBLiTSicMA, A. F. 1981. Opredelitel" molodi presnovodnykh ryb.
Legkaia i pishevaia promyshlennost", Moskova.
KocK, K. H. 1981. Fischereibiologische Untersuchungen an drei ant-
arktischen Fischarien: Champsocephaius gunnari Lonnberg, 1905,
Chaenocephalus aceratus (Lonnberg, 1906) und Pseudochaen-
ichlhys georgianus Norman, 1937 (Notothenioidei, Channichthyi-
dae). Mitt. Inst. Seefisch. 32:1-226.
KoEFOED, E. 1909. Poissons. Croisiere oceanographique accompli a'
bord de la Belgica dans la mer de Gronland 1905. Bruxelles.
. 1927. Fishes from the sea-bottom. Rep. Sci. Results Michael
Sars N. Atl. Deep-Sea Exped. 4(Pt. 1): 1-147.
. 1953. Synentognathi, Solenichthyes, Anacanthini, Beryco-
morphi, Xenoberyces. Rep. Sci. Results Michael Sars N. Atl. Deep-
Sea Exped. 4(Pt. 2, No. 3):3-36.
KoENiG, C. C, AND R. J. Livingston. 1976. The embryological de-
velopment of the diamond killifish (Adinia xentca). Copeia 1976:
435-445.
KoHNO, H., T. HosHiNO, F. Yasiida and Y. Taki. 1982. Larval me-
lanophore patterns of Thunnus alatunga and T. thynmis from the
Mediterranean. Jpn. J. Ichthyol. 28:461-465.
K0K.HANOVA, N. A. 1957. Razvitie shchipovki (Cohitis taenia L.).
Vopr. Ikhtiol. 8:89-101.
KoMAROv, Y. A. 1964. Some data of the reproduction of the horse
mackerel- rrac/jwr;/.? trachurus L. from the Southwest African coast.
Tr. Atl. Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. 1 1:87-99.
Kopec, J. A. 1 949. Ecology, breeding habits and young stages of Cren-
ichthys baileyi. a cyprinodont fish of Nevada. Copeia 1949:56-61.
KORMENDY. A. C. 1975. Microorganisms, p. 82-108. In: Principles
and techniques of scanning electron microscopy. Vol. 3. M. A.
Hayat (ed.). Van Nostrand Reinhold Co.. New York.
KosTER, W. J. 1948. Notes on the spawning activities and the young
slzges of Plancterus kansae (GiTman). Copeia 1948:25-33.
KoTLYAR, A. N. 1981. A new family, genus and species of Beryci-
formes, Hispidoberycidae fam., Hispidoberyx ambagiosus gen. et
sp. n. (Beryciformes). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
21(3):9-13.
KoTTHAUs, A. 1969. 2. FischedesIndischenOzeans. A. Systematisch-
er Teil, V. Solenichthyes und Anacanthini. Meteor Forschungser-
geb. Reihe D. Biol. 4:32-46.
. 1970. Flagelloserranus a new genus of serranid fishes with the
LITERATURE CITED
715
description of two new species (Pisces, Percomorphi). Dana-Rep.
Carlsberg Found. 78.
. 1972. Fische des Indischen Ozeans. A. Systematischer Teil,
IX. Iniomi (Nachtrag: Fam. Myctophidae). Meteor Forschungser-
geb. Reihe D Biol. 12:12-35.
KouMANS, F. P. 193 1 . A preliminary revision of the genera of gobioid
fishes with united ventral fins. Drukkerij Imperator. N.V.. Lisse.
Kdvalevskava, N. V. 1964. Study of development of the "two winged"
flying fishes of the genus Exocoelus (Exocoelidae, Pisces). Tr. Inst.
Okeanol. Akad. Nauk SSSR 73:204-223. (In Russian, English sum-
mary.)
. 1965. The eggs and larvae of synentognathous fishes (Belon-
iformes, Pisces) of the Gulf of Tonkin. Tr. Inst. Okeanol. Akad.
Nauk. SSSR 80:124-146. (In Russian, English summary.)
. 1972. New materials on reproduction, development and dis-
tribution of larvae and juveniles of flying fishes of the genus Hi-
rundichthys (Exocoetus, Pisces) in the Pacific and Indian Oceans.
Tr. Inst. Okeanol. Akad. Nauk SSSR 93:42-69. (In Russian, English
summary.)
. 1982. Superfluous reproduction and development of flying
fishes of the family Exocoetidae. J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 22(4):48-54.
KowTAL, G. V, 1972. Observations on the breeding and larval de-
velopment of Chilka 'Sabal' Eleutheronema telradeclyturn (Shaw).
India J. Fish. 19(l-2):70-75.
KoZHOV, M. 1963. Lake Baikal and its life. Monographiae Biologicae,
Vol. II. Dr. W. Junk. The Hague.
KozLOV, B. M. 1952. Observations on the development of navaga
eggs. Izv. Tikhookean Nauchno-Issled. Inst. Rybn. Khoz. Okean-
ogr. 34:261-262. (In Russian.)
Kramer, D. 1960. Development ofeggs and larvae of Pacific mackerel
and distribution and abundance of larvae 1 952-56. U.S. Fish Wildl.
Ser. Fish. Bull. 60:393-438.
Kraul, S. 1983. Results and hypotheses for the propagation of the
grey mullet, MugU cephatus L. Aquaculture 30:273-284.
Krefft, G. 1 96 1 . a contnbution to the reproductive biology of Hel-
icolenus daclylopterus (De la Roche, 1809) with remarks on the
evolution of the Sebaslinae. Rapp. P. V. Reun. Cons. Perm. Int.
Explor. Mer 150:243-244.
. 1970. Zur Systematik und Verbreitung der Gattung Lampa-
dena Goode and Bean. 1896 (Osteichthyes, Myctophoidei, Myc-
tophidae) im Atlantischen Ozean, mit Beschreibung einer neen Art.
Ber. Dtsch. Wiss. Komm. Meeresforsch. 21:271-289.
. 1974. Investigations on midwater fish in the Atlantic Ocean.
Ber. Dtsch. Wiss. Komm. Meeresforsch. 23:226-254.
Kryzhanovskii, S. G. 1947. Sistema semeistva karpovykh ryb (Cy-
prinidae). Zool. Zh. 26:53-64.
. 1948. Ekologicheskie gruppy ryb i zakonomemosti ikh raz-
vitiia. Izv, Tikhookean Nauchno-Issled. Insl. Rybn. Khoz. Okean-
ogr. 27:5-114.
. 1949. Ekologo-morphologicheskie zakonomemosti razvitiia
karpovykh, v'iunoviykh i somovykh ryb(Cyprinoidei i Siluroidei).
Tr. Inst. Morfol. Zhivotn., Akad. Nauk SSSR 1:1-232.
, A. I. Smirnov and S. G. Soin. 1951. Materialy po razvitiiu
ryb r. Amura. Truy Amurskoi Ikhtiologicheskoi Ekspeditsii 1945-
1949. Mater. Poznaniyu Fauny Flory SSSR otd. Zool. 24:5-222.
KuBO, I. 1939. Note on the development of a teleost, Thalassoma
cupido (Temminck and Schlegel). Bull. Jpn. Soc. Sci. Fish. 8(3):
165-167. (In Japanese. English summary.)
KucHNOw, K. P., andJ. R. Scott. 1977. Ultrastruclure of the chorion
and its micropyle apparatus in the mature Fundulus heleroclitus
ovum. J. Fish Biol. 10:197-201.
KuMLMAN, D., G. Qliantz AND U. WiTT. 1981. Rearing of turbot
larvae {Scophlhahnus ma.ximus L.) on cultured food organisms and
postmetamorphosis growth on natural and artificial food. Aqua-
culture 23:183-196.
KiiLKARNi, C, V. 1940. On the systematic position, structural modi-
fications, bionomics and development of a cyprinodont fishes from
the province of Bombay. Rec. Indian Mus. (Calcutta) 42:379-423.
. 1950. Breeding habits, eggs and early life history of the Indian
shad. HUsa ilisha (Ham.), in the Narbada River, Proc. Natl. Inst,
Sci. (India) 16:169-176.
Kuntz, A. 1914a. Noteson the habits, morphology of the reproductive
organs and embryology of the viviparous fish Gambusia affinis.
Bull. Bur. Fish. Wash. 33(1913):181-I90.
. 1 9 1 4b. The embryology and larval development of Bairdiella
chrysura and Anchovia mitchilli. Bull. Bur. Fish. Wash. 33:1-19.
. 1916. Notes on the embryology and larval development of
five species of teleosteon fishes. Bull. Bur. Fish. Wash. 34(1914):
407-429.
, AND L. Radcliffe. 1917. Notes on the embryology and larval
development of twelve teleostean fishes. Bull. Bur. Fish. Wash. 35:
87-134.
KuNZ, Y. 1964. Morphologische Sludien iiber die embryonale und
postembryonal Entwicklung bei Teleostiem mit besonderer Be-
rijcksichtigung des Dotlersystems und der Leber. Rev. Suisse Zool.
71:445-525.
Kro. C-M.. H. Shehadeh and K. K. Milisen. 1973. A preliminary
report on the development, growth and survival of laboratory reared
larvae of the grey mullet, Mugil cephalus L. J. Fish Biol. 5:459-
470.
Kurata, H. 1956. On the rearing of larvae of the flatfish, Liopsetta
obscura in small aquaria. Bull. Hokkaido Reg. Fish. Res. Lab., Fish.
Agency 13:20-28.
Kuroda, K.,T. Yamamoto and Y. Hirano. 1982. Development and
identification of the egg of the Pacific mackerel (Scomber japonicus
Houttuyn) and a similar egg. Bull. Tokai Reg. Fish. Res. 107:33-
52.
KuTHALiNGAM, M. D. K. 1959a. A contribution to the life histories
and feeding habits of horse-mackerels, Megalaspis cordyla (Linn.)
and Caranx mate (Cuv. & Val.) and notes on the effect of absence
of light on the development and feeding habits of larvae and po%X-
larvae of Megalaspis cordvla. J. Madras Univ. Sect. B 29(2):79-
96.
. 1959b. The life history and feeding habits of a tripod fish
( Tnacanlhus breviroslns Temm. & Schleg.) of the Indian Sea. Treu-
bia 25:159-164.
KuwABARA, S. 1954. Occurrence of luminous organs on the tongue
of two scopelid fishes, Neoscopeliis macrolepidotus and N. micro-
chir. J. Schimonoseki Coll. Fish. 3:283-287.
Kuz'min-Karovaev. G. A. 1930. Biology of the White Sea navaga.
Tr. Nauchno. Inst. Rybn. Khoz. (5) 4:83-86. (In Russian.)
Kvenseth, P. G., AND V. Oiestad. 1984. Large-scale rearing of cod
fry on the natural food production of an enclosed pond, p. 644-
655. In: The propagation of cod. E. Dahl, D. Danielssen, E. Moks-
ness and P. Solemdal (eds.). Flodevigen Rapp. I.
Kyi.e, H. M. 1913. Flat-fishes (Helerosomata). Rep. Dan. Oceanogr.
Exped. Mediterr. 2(A.I):1-150.
. 1921. The asymmetry, metamorphosis and origin of flat-fish-
es. Philos. Trans. R. Soc. London Ser. B 21 1:75-129.
KvusHiN, K. 1968. The embryonic and larval stages of //ew/rnpferwj
v///oj«i (Pallas). Bull. Fac. Fish. Hokkaido Univ. 18:277-289.
. 1970. Embryonic development and larvae of G.vmnoca«?/!!«
herzensteini JovAan and Starks. Jpn. J. Ichthyol. 17:74-79.
1975. The embryonic and larval development, growth, sur-
vival and changes in body form, and the effect of temperature on
these characteristics of the smooth lumpsucker. Aplocychis ventn-
cosus (Pallas). Bull. Fac. Fish. Hokkaido Univ. 26(l):49-72.
Laale, H. W. 1980. The perivitellinc space and egg envelopes of bony
fishes: a review, Copeia 1980:210-226.
Laciardere, F. 1980. Developpement du ceteau, Dicologoglossa cu-
neata (Moreau) (Poissons-Soleidae). Cybium, 3rd Ser. 4(4):61-81.
, and a. Aboussohan. 1981. Developpement du ceteau, Di-
cologoglossa cuneala (Moreau, 1881) (Pisces, Pleuronectiformes,
Soleidae): II — Description des larves. Cybium, 3rd Ser. 5(2):53-72.
716
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Lai, J. 1 979. The development of stages of mullet, Mugil so-nig Bas-
ilewsky L. Acta. Oceanol. Sin. 1(1): 157-1 75.
Lake, J. S. 1967. Rearing experiments with five species of Australian
freshwater fishes. II. Morphogenesis and ontogeny. Aust. J. Mar.
Freshwater Res. 18:155-173.
Lam, T. J. 1982. Applications of endocnnology to fish culture. Can.
J. Fish. Aquat. Sci. 39:111-137.
Laroche, J. L., S. Richardson AND A. A. Rosenberg. 1982. Age and
growth of a pleuroneclid, Parophrys velulus. during the pelagic lar-
val period in Oregon coastal waters. Fish. Bull. U.S. 80:93-104.
Laroche, W. A. 1977. Descnplion of larval and early juvenile ver-
milion snapper, Rhomboplites aurorubens. Fish. Bull. U.S. 75:547-
554.
. 1981. Development of larval smooth flounder, Z.;opieKa/)Mr-
nami. with a redescription of winter flounder, Pseudopleuronectes
amencanus (Family Pleuronectidae). Fish. Bull. U.S. 78:897-909.
, AND S. L. Richardson. 1980. Development and occurrence
of larvae and juveniles of the rockfishes Sebasles JIavidits and Se-
bastes melanops (Scorpaenidae) off Oregon. Fish. Bull. U.S. 77:
901-924.
, and . 1981. Development of larvae and juveniles of
the rockfishes Sebasles enlomelas and 5. zacemrus (Family Scor-
paenidae) and occurrence ofli' Oregon, with notes on head spines of
S. mrslinus. S. flavidus and S. melanops. Fish. Bull. U.S. 79:231-
257.'
, D. Rupee and S. L. Richardson. MS. Young carangid fishes
from the Gulf of Mexico: A generic diagnosis of larvae, with ad-
ditional descriptive notes including early development of three
species. Gulf Coast Research Lab., Ocean Springs, MS 39564.
L.\siv£R, R. 1975. Field criteria for survival of anchovy larvae: the
relation between inshore chlorophyll maximum layers and suc-
cessful first feeding. Fish. Bull. U.S. 73:453-462.
. 1981. The role of a stable ocean in larval fish survival and
subsequent recruitment, p. 80-87. In: Marine fish larvae. Mor-
phology, ecology and relation to fishenes. R. Lasker (ed.). Univer-
sity of Washington Press, Seattle, WA.
, H. M. Feder, G. H. TheilackerandR.C. May. 1970. Feed-
ing, growth and survival of Engraulis mordax reared in the labo-
ratory. Mar. Biol. 5:345-353.
, AND K. Sherman (eds.). 1981. The early life history offish:
recent studies. Rapp. P.V. Reun. Cons. Int. Explor. Mer 178:1-
607.
, and L. T. Threadgold. 1968. "Chloride cells" in the skin of
larval sardine (Sardinops caendea). Exp. Cell Res. 52:582-590.
Lau, S. R., and p. L. Shafland. 1982. Larval development of the
snook, Ceniropomus undecimalis (Pisces: Centropomidae). Copeia
1982:618-626.
Lauder, G. V., Jr. 1980. On the relationship of the myotome to the
axial skeleton in vertebrate evolution. Paleobiology 6:51-56.
, AND K. F. LiEM. 1983. The evolution and interrelationships
of the Actinopterygian fishes. Bull. Museum Comp. Zool. Harv.
Univ. 150(31:95-197.
Laurence, G. C. 1981. Overview — modelling— an esoteric or poten-
tially utilitarian approach to understanding larval fish dynamics.
Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:3-6.
Lea, R, N. 1980. Systematics and zoogeography of cusk-eels of the
family Ophidiidae, subfamily Ophidiinae. from the eastern Pacific
Ocean. Unpubl. PhD dissertation. Univ. Miami, Coral Gables, FL.
Leach, G. C. 1923. Artificial propagation of whitefish. grayling and
lake trout. Rep. U.S. Comm. Fish. Appendix 3, Doc. 949:1-32.
Lebour, M. V. 1919. The young of the Gobiidae from the neighbor-
hood of Plymouth. J. Mar. Biol. Assoc. U.K. 12:48-80.
. 1921. The larval and post-larval stages of the pilchard, sprat,
and herring from the Plymouth District. J. Mar. Biol. Assoc. U. K.
12:427-457.
. 1927. The eggs and newly hatched young of the common
blennies from the Plymouth neighbourhood. J. Mar. Biol. Assoc.
U.K. 14:647-650.
Le Danois, Y. 1959. Etude osteologique, myologique et systematique
des poissons du sous— ordre des Orbiculales. Ann. Inst. Oceanogr.
36(1).
. 1961. Remarques sur les poissons orbiculates du sous-ordre ^
des Ostracioniformes. Mem. Mus. Natl. Hist. Nat. Ser. A Zool. 29:
207-338.
Legand, M. 1967. Cycles biologiques des poissons mesopelagiques
dans Test de I'ocean Indian. Premiere note. Scopelopsis midti-
punclalus Brauer, Gonostoma sp.. Notolychnus valdiviae Brauer.
Cah ORSTOM Ser. Oceanogr. 5(4):47-7l.
Leibi, M. M. 1979a. Leptocephalus larvae of the eel family Ophichth-
idae. I. Ophichlhus gomesi Ca^le\nau. Bull. Mar. Sci. 29:329-343.
. 1979b. Morphological development of the eel Myrophis punc-
tatus (Ophichthidae) from hatching to metamorphosis, with em-
phasis on the developing head skeleton. Bull. Mar. Sci. 29:509-
521.
. 1981. Larval morphology of the eels 5a«'a«;t7!?/;i'i /la^ca/J-
ium. B scuticaris. Ophichthus melanoporus and O. ophis
(Ophichthidae), with a discussion of larval identification methods.
Bull. Mar. Sci. 31:46-71.
. 1982. Leptocephalus larvae ofthetnbeSphagebranchini (Pisces.
Ophichthidae) in the western North Atlantic. Bull. Mar. Sci. 32:
220-236.
. 1984. Leptocephalus larvae of the tribe Callechelyini (An-
guilliformes, Ophichthidae, Ophichthinae) in the western North
Atlantic. Bull. Mar. Sci. 34:398^23.
. In press. Ophichthidae. /«. Leptocephalus larvae of the Gulf
of Mexico and Yucatan Channel. D. G. Smith and W. H. Hulet
(eds.). Stud. Trop. Oceanogr. (Miami).
Leim, a. H., and W. B. Scott. 1966. Fishes of the Atlantic coast of
Canada. Bull. Fish. Res. Board Can. 155.
Leis, J. M. 1977. Development of the eggs and larvae of the slender
mola, Ranzania laevis (Pisces, Molidae). Bull. Mar. Sci. 27:448-
466.
. 1978. Systematics and zoogeography of the porcupine fishes
{Diodon. Diodontidae, Tetraodonliformes), with comments on egg
and larval development. Fish. Bull. LI.S. 76:535-567.
. 1982. Hawaiian creediid fishes (Cryslallodytes cookei and
Limnichthys donaldsoni): development of eggs and larvae and use
of pelagic eggs to trace coastal water movement. Bull. Mar. Sci. 32:
166-180.
. 1983. Coral reef fish larvae (Labndae) in the East Pacific
Barrier. Copeia 1983:826-828.
, andB. GC5LDMAN. 1983. Studies On the biology of larval fishcs
in the Lizard Island area. Northern Great Barrier Reef, p. 22 1 -22 5 .
In: Proc. Inaugural Great Barrier Reef Conf i. T. Baker. R. M.
Carter, P. W. Sammarco and K. P. Stark (eds.). J. Cook University
Press, Townsville, Australia.
, AND J. T. Mover. MS. Development of eggs and larvae of
three Indo-Pacific trunkfishes (Tetraodonti formes, Ostraciidae).
, AND D. S. Rennis. 1983. The larvae of Indo-Pacific coral reef
fishes. New South Wales University Press, Sydney.
Lentz, T. L.. AND J. P. Trinkhaus. 1967. A fine structural study of
cytodifferentiation during cleavage, blastula. and gastrula stages of
Fundidus hcteroditus. J. Cell. Biol. 32:121-138.
Leonard, S. B. 1971. Larvae of the fourspot flounder, Hippoglossina
oblonga (Pisces: Bothidae), from the Chesapeake Bight, western
North Atlantic. Copeia 1971:676-681.
LEONci, R. 1971. Induced spawning of the northern anchovy Engraulis
mordax Girard. Fish. Bull. U.S. 69:357-360.
. 1977. Maturation and induced spawning of captive Pacific
mackerel. Scomber japomcus. Fish. Bull. U.S. 75:205-21 I.
Letaconnoux, R. 1951. Contribution a I'etude des especes du genre
Trachuriis et specialement du Trachurus irachiirus (L.). Mem. Off.
Sci. Pech. Mant. 15:1-67.
Levings, C. D. 1 969. The zoarcid Lycodopsis pacifica in outer Burtard
Inlet, British Columbia. J. Fish. Res. Board Can. 26:2403-2412.
LlAO, I. C, J. C. JUARIO, S. KUMACiAI, H. Nakajima. M. Natividad
and p. Buri. 1979. On the induced spawning and larval rearing
of the milkfish, Chanos chanos (Forskal). Aquaculture 18:75-93.
LITERATURE CITED
717
LiMBAUGH, C. 1962. Life history and ecological notes on the tubenose,
Aulorhynchus IJavidus. a hemibranch fish of western North Amer-
ica. Copeia 1962:549-555.
LiNDBERG, G. U., AND Z. V. Krasyukova. 1975. Fishcs of the Sea of
Japan and the adjacent areas of the Sea of Okhotsk and the Yellow
Sea. Part 4. Nauka. Leningrad. (In Russian.)
. and M. I. Legeza. 1955. A review of the genera and species
of the subfamily Cyclopterinae (Pisces). Bull. Zool. Inst. Acad. Sci.
USSR 18:389-458.
, AND . 1969. Fishes of the Sea of Japan and the adjacent
areas of the Sea of Okhotsk and the Yellow Sea. Part 2. Teleostomi.
Xll. Acipenseriformes — XXVlll. Polynemiformes. Akad. Nauk.
SSSR. No. 84: 1-389. (First published in 1 965. Transl. from Russian
by Isr. Prog. Sci. Transl. cat. no. 1852. Available from: NTIS,
Springfield, VA; TT68-50361.)
LiNDSEV, C. C. 1975. Pleomerism, the widespread tendency among
related fish species for vertebral number to be correlated with max-
imum body length. J. Fish. Res. Board Can. 32:2453-2469.
LiPPSON, A. J., AND R. L. MoRAN. 1974. Manual for identification of
early developmental stages of fishes of the Potomac River Estuary.
Md. Dep. Nat. Resour. Power Plant Siting Program PPSP-MP-13:
236-241.
LisovENKO, L. A., andM. F. SvETLOV. 1981. Some data on the biology
of Paraliparis (Family Liparidae) off the South Shetlands. J. Ich-
thyol. (Engl. Transl. Vopr. Ikhtiol.) 20(1):131-135.
, A. V. Barabanov AND V. N. Efremenko. 1982. New data on
the reproduction of the "southern putassu". Micromesistius aus-
tralis (Gadidae), from the Falkland-Patagonian zoogeographical re-
gion. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 22(4):55-67.
Llewellyn, L. C 1973. Spawning, development, and temperature
tolerence of the spangled perch, Madigania unicolor (Gunther),
from inland waters in Australia. Aust. J. Mar. Freshwater Res. 24:
73-94.
. 1974. Spawning, development and distribution of the southern
pigmy perch Nannoperca australis australis Gunther from inland
waters of eastern Australia. Aust. J. Mar. Freshwater Res. 25:121-
149.
LOBEL, P., AND R. Johannes. 1980. Nesting, eggs and larvae of trig-
gerfishes (Balistidae). Environ. Biol. Fishes 5:251-252.
Lo Bianco, S. 1903. Le pesche abissali esequite da F. A. Krupp col
yacht (Puritan) nelle adiacenze di Capri ed in altre localita del
Mediterraneo. Mitt. Zool. Sta. Neapel 16:109-279.
. 1908a. Uova e larve di Trachipterus taenia Bl. Mitt. Zool.
Sta. Neapel. 19(1): 1-1 7.
. 1908b. Sviluppo larvale, metamorfos! biologia della "Tnglia
di fango" {.\fullus harhatus Lin.) Mitt. Zool. Stn. Neapel 19:18.
. 1909. Notize biologiche riguardanti specialmente il periodo
di maturita sessuale delgi animali del Golfo di Napoli. Mitt. Zool.
Sta. Neapel 19(4):5 13-763.
. 1911. Su alcuni stadi postlarvali appartenenti a Gadidi ran
del golfo di Napoli. Mitt. Zool. Sta. Neapel 20:170-187.
LofKET, N. A. 1970. Eyes for the deep. Spectrum 78:10-15.
Loc KwooD. S. J.. J. H. Nkhols and W. A. Dawson. 1981. The
estimation of a mackerel {Scomber scomhrus L.) spawning stock
size by plankton survey. J. Plankton Res. 3:217-233.
LoEB, V. J. 1979. Larval fishes in the zooplankton community of the
North Pacific central gyre. Mar. Biol. 53:173-191.
LoNOHiRST, A. R. 1971. The clupeoid resources of tropical seas.
Oceanogr. Mar. Biol. Annu. Rev. 9:349-386.
LoNNiNii, S. 1972. Comparative electron microscopic studies of te-
leostean eggs with special reference to the chorion. Sarsia 49:41-
48.
, AND B. E. Hagstrom. 1975. Scanning electron microscope
studies of the surface of the fish egg. Astarte 8:17-22.
Loos, J. J., AND L. A. Fliiman. 1978. Subordinate taxa of the genus
Nolropis: a preliminary survey of their developmental traits, p. 1-
50. In: Proc. First Symp. on Freshwater Larval Fish. L. L. Olmsted
(ed.). Southeastern Electric Exchange.
, , E. K. Jankowski AND N. R. Foster. 1979.
Notes on early life histories of cyprinoid fishes of the upper Potomac
River, p. 93-139. In: Proceedings of a workshop on freshwater
larval fishes. R. Wallus and C. W. Voigtlander (eds). Tennessee
Valley Authority. Noms, TN.
Lopez RojAS, H., AND A. Machado Allison. 1975. Algunos aspectos
del desarrollo y crecimiento de Loricaria laticeps (Osteichthyes,
Siluriformes, Loricaridae). Acta Biol. Venez. 9(I):51-76.
Lough, R. G., M. Pennington, G. R. Bolz and A. A. Rosenberg.
1982. Age and growth of larval Atlantic herring, Clupea harengus
L,, in the Gulf of Maine-Georges Bank region based on otolith
growth increments. Fish. Bull. U.S. 80:187-199.
Louw, E., and M. J. OTooLE. 1977. Larval development ofSardinops
OfWtoa (Pisces: Clupeidae). Ann. S. Afr. Mus. 72(7):125-145.
Lowell, N. E. 1971. Some aspects of the life history and spawning of
the moi (Polydaclylus saxtitis). Unpubl. MS thesis. Univ. Hawaii,
Honolulu.
LowENSTAM, H. A. 1980. Bioinorganic constituents of hard parts, p.
3-16. In: Biogeochemistry of Amino Acids. P. E. Hare (ed). John
Wiley & Sons, New York.
. 1981. Minerals formed by organisms. Science21 1:1 126-1 131.
, AND J. E. Fitch. 1978. Morphology, mineral diversity and
ontogenetic mineralogic changes in extant elasmobranch otoconia.
Geol. Soc. Am. Abstr. 10:1 14.
, and . 1981. Vatente formation in otoliths of higher
teleost fishes. Abstract. Pacific Division. AAAS, Annual Meeting,
June 14-19, 1981, Univ. of Oregon. Eugene.
Llibbock, R. 1975. Fishes of the family Pseudochromidae (Perci-
formes) in the northwest Indian Ocean and Red Sea. J. Zool. (Lond.)
176(2):115-157.
Lund, W. A., Jr., AND B. C Mar< I, Jr. 1975. Early development of
the grubby, Myoxocephalus aeneas (Mitchell). Biol. Bull. (Woods
Hole) 149:373-383.
Lundberg, J. G. 1972. Wagner networks and ancestors. Syst. Zool.
21:398-413.
Llither, G. 1975a. New characters for consideration in the taxonomic
appraisal of grey mullets. Aquaculture 5:107. (Abstract.)
. 1975b. On the little-known mullet, I'alamugil buchanani
(Bleeker) (Mugilidae: Pisces) in the Indian waters. J. Mar. Biol.
Assoc. India 16(1):290-291.
LuTKEN, C. 1880. Spolia Atlantica. Bidrag til kundaskab on forafor-
andringer nos fiske under deres vaext og udvikling. saerlight nus
nogle of Atlantic Lavets honosfiske. Kgt. Dan. Vidensk. Selsk. Biol.
Skr. 5{12):409-613. (In Danish. French summar>.)
Lythgoe, J., AND G. Lythgoe. 1975. Fishes of the sea. Anchor Press/
Doubleday, Garden City, New York.
Maccall, A. D. 1980. Theconsequencesof cannibalism in the stock-
recruitment relationship of planktivorous pelagic fishes such as
Engraulis. IOC Workshop Rep. No. 28:201-220.
MacDonald, C. M. 1978. Morphological and biochemical system-
atics of Australian freshwater and estuarine percichthyid fishes.
Aust. J. Mar. Freshwater Res. 1978:667-698.
Macer, C. T. 1967. Ammodytidae. Cons. Int. Explor. Mer, Fiches
Identification Oeufs Larves Poissons. 2.
MacGregor, J. S. 1976. Ovarian development and fecundity of five
species of California current fishes. Calif Coop. Oceanic Fish. In-
vest. Rep. 18:181-188.
M^^HADO Allison, A., and H. Lopez Rojas. 1975. Etapas del de-
sarrollo de Loncanichthys lypiis (Bleeker) 1864 (Osteichthyes, Si-
luriformes, Loncanidae). Acta Biol. Venez. 9(1):93-1 19.
Major. P. F. 1973. Scale feeding behaviour of the leatherjacket. Scom-
beroides lysan (Forskal) and two species of the genus OligopHles
(Pisces: Carangidae). Copeia 1973:151-154.
Makeeva, a. p. 1 976. Characteristics of the early de\ elopmental stages
of a bitterling Rhodeus ocellalus occllalus new to the ichthyofauna
of the USSR. J. Ichthyol. (English Trans. Vopr. Ikhtiol.) 16:756-
767.
, and I. N. Ryabov. 1973. Reproduction and development of
Opsariichthys uncirostris (Temminck et Schlegel) and its systematic
718
ONTOGEl^fY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
position in the family Cyprinidae. J. Ichthyol. (English Transl. Vopr.
Ikhtiol.) 13:392-399.
Maksimov, V. P. 1969. Onlhe spav/ning of Naiicralesducloi: Tr. Atl.
Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. 25:206-209. (In Rus-
sian, taken from Biol. Abstr., 1971, abstr. no. 42506.)
Makushok, V. M. 1958. The morphology and classification of the
northern blennioid fishes (Stichaeoidae, Blennioidei. Pisces). Tr.
Zool. Inst. Akad. Nauk SSSR 25:3-129. (Transl. by A. R. Gosline,
1959, Ichthyol. Lab.. U.S. Fish Wildl. Serv., U.S. Natl. Mus., Wash..
D.C.)
. 1961. The group Neozoarcmae and its place in the classifi-
cation of fishes (Zoarcidae. Blennioidei, Pisces). Tr. Zool. Inst. Akad.
Nauk SSSR 43:198-224.
. 1 966. The specific identity of Nematonums longifilis (Gunther.
1877) and N. clarki (Jordan and Gilbert, 1898) and some remarks
on age-dependant variations in Macruridae (Pisces), p. 147-172.
In: Fishes of the Pacific and Indian Oceans. Biology and Distri-
bution. T. S. Rass (ed.). Tr. Inst. Okeanol. Akad. Nauk. SSSR 73:
147-172.
Malloy, R., and F. D. Martin. 1982. Comparative development of
redfin pickerel {Esox americanus americanus) and the eastern mud-
minnow (Umbra pygmaea), p. 70-72. In: Proc. 5th Annual Larval
Fish Conference. C. F. Bryan, J. V. Conner and F. M. Truesdale
(eds.). Louisiana Cooperative Fishery Unit, L.S.U., Baton Rouge,
LA.
Manacop. p. R. 1936. A new phallostethid fish with notes on its
development. Philipp. J. Sci. 59:375-381.
Mancuso, B. 1926. Contributo alia conoscenza dello svilupo larvale
di Physiculus dalwigki Kp. Mem. R. Comm. Tallasogr. Ital. 124.
Mansueti, a. J. 1964. Early development of the yellow perch, Perca
flavescens. Chesapeake Sci. 5(I-2):46-66.
, AND J. D. Hard>. 1967. Development of fishes of the Ches-
apeake Bay region: An atlas of egg, larval, and juvenile stages. Part
I. Nat. Resour. Inst. Univ. MD Port City Press. Baltimore. MD.
Mansueti, R, J. 1958. Eggs, larvae, and young of the striped bass,
Roccus saxalilis. Md. Dep. Res. Educ. CBL Contrib. 1 12.
. 1963. Symbiotic behavior between small fishes and jellyfishes
with new data on that between the stromateid, Pepnlus atepidotus.
and the Scyphomedusa, Chrvsaora qumquecirrha. Copeia 1 963:40-
80.
. 1964. Eggs, larvae, and young of the white perch, Roccus
americanus. with comments on its ecology in the estuary. Chesa-
peake Sci. 5(l-2):3-45.
Marak, R. R. 1967. Eggs and early larval stages of the offshore hake,
Merluccius albidus. Trans. Am. Fish. Soc. 96:227-228.
Marathe, V. B., AND D. V. Bal. 1958. A brief comparative account
of the axial skeleton of six polynemids from Bombay waters. J.
Univ. Bombay 268:139-151.
Marchal, E. G. 1966. Oeufs, larves et post-larves de Tanchois du
Golfe de Guinee, Anchoviella guineensis(Q\ac]:\c et Rossignol). Doc.
Sci. Prov. (Abidjan) No. 005SR.
. 1967. Cle provisoire de determination des oeufs et larves des
Clupeidaeet Engraulidae Ouest-africains. Doc. Sci. Prov. (Abidjan)
No. 0I4SR.
Marinaro. J.-Y. 1971. Contribution a I'etude des oeufs et larves
pelagiques de poissons mediterraneens. V. Oeufs pelagiques de la
bale d' Alger. Pelagos 3(l):l-l 15.
Marino, R. P., AND J. K. Dooley. 1982. Phylogenetic relationships
of the tilefish family Branchiostegidae (Perciformes) based on com-
parative myology. J. Zool. (Lond.) 1 982(1 96): 15 1-1 63.
Marion, A. F. 1894a. Sur la peche et la reproduction du "Sioudet"
(Atherina hepsetus L.). Ann. Mus. Hist. Nat. Marseille (Zool.)
4:93-99.
. 1894b. Oeufs flotlants et alevins observes dans le Golfe de
Marseille durant I'annee 1 890. Ann. Mus. Hist. Nat. Marseille 4( 1 1 ):
112-121.
Markle, D. F. 1982. Identification of larval and juvenile Canadian
Atlantic gadoids with comments on the systematics of gadid
subfamilies. Can. J. Zool. 60(12):3420-3438.
— . 1984. Phosphate buffered formalin for long term preservation
of formalin fixed ichthyoplankton. Copeia 1984:525-528.
— , AND L.-A. Frost. MS. Comparative morphology and iden-
tification of planktonic fish eggs from the Nova Scotian Shelf
— , AND J. E. Olney. 1980. A description of the vexillifer larvae
of Pyramodon venlralis and Snydendia canma (Pisces, Carpaidae)
with comments on classification. Pac. Sci. 34:173-180.
— , W. B. Scott AND A. C. KoHLER. 1980. New and rare records
of Canadian fishes and the influence of hydrography on resident
and nonresident Scotian Shelf ichthyofauna. Can. J. Fish. Aquat.
Sci. 37:49-65.
— , AND C. A. Wenner. 1979. Evidence of demersal spawning in
the mesopelagic zoarcid fish Melanostigma allanlicum. with com-
ments on demersal spawning in the alepocephalid fish Xenoder-
michlhys copei. Copeia 1979:363-366.
, J. T. Williams AND J. E. Olney. 1983. Description of a new
species of Echiodon (Teleostei, Carapidae) from Antarctic and ad-
jacent seas. Proc. Biol. Soc. Wash. 97(1).
Marliave, J. B. 1975. The behavioral transformation from the plank-
tonic larval stage of some marine fishes reared in the laboratory.
Unpubl. PhD dissertation. Univ. British Columbia, Vancouver.
. 1976. A theory of storm-induced drift dispersal of the gas-
terosteid fish Aulorhynchus Jlavidus. Copeia 1976:794-796.
. 1978. Spawning and yolk-sac larvae of the agonid fish species.
Agonomalus mozinoi Wilimovsky and Wilson. Syesis 1 1:285-286.
. 1981. Spawn and larvae of the Pacific sandfish. Trichodon
tnchodon. Fish. Bull. U.S. 78:959-964.
Marshall, N. B. 1953. Egg sizes of Arctic, Antarctic and deep-sea
fishes. Evolution 7:328-341.
. 1955. Alepisauroid fishes. Discovery Rep. 27:303-336.
. I960. Swimbladder structure of deep-sea fishes in relation to
their systematics and biology. Discovery Rep. 31:1-122.
. 1961. A young Macnslium and the ctenothrissid fishes. Bull.
Br. Mus. (Nat. Hist.) Zool. 7:355-370.
. 1962. Observations on the Heteromi, an order of teleost fishes.
Bull. Br. Mus. (Nat. Hist.) Zool. 9:249-270.
. 1964. Bathypelagic macround fishes. Copeia 1964:86-93.
. 1965. Systematic and biological studies oflhemacrourid fishes
(Anacanthini-Teleostii). Deep-Sea Res. 12:299-322.
. 1966a. Family Scopelosauridae. p. 194-204. />;.■ Fishes of the
western North Atlantic. D. M. Cohen (ed.). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 5).
. 1966b. The relationships of the anacanthine fishes. Macru-
ronus, Lyconus. and Steiridachnena. Copeia 1966:275-280.
— . 1971. Explorations in the life of fishes. Harvard Univ. Press,
Cambridge. MA.
— . 1973. Family Macouridae. p. 496-665. /«. Fishes of the west-
em North Atlantic. D. M. Cohen (ed.). Mem. Sears Found. Mar.
Res. No. 1 (Pt. 6).
— , AND D. M. Cohen.
1973. Order Anacanthini (Gadiformes).
Characters and synopsis of families, p. 479-495. In: Fishes of the
western North Atlantic. D. M. Cohen (ed.). Mem. Sears Found.
Mar. Res. No. I (Pt. 6).
, AND T. IwAMOTO. 1973. Family Macrouridae (Genera Coe-
lorhynchus. Coryphaenoides, Hymenocephahis and Nczumia). p.
496-665. In: Fishes of the western North Altantic. D. M. Cohen
(ed.). Mem. Sears Found. Mar. Res. No. I (Pt. 6).
, and J. C. Staiger. 1975. Aspects of the structure, relation-
ships, and biology of the deep-sea fish Ipnops murrayi (Family
Bathypteroidae). Bull. Mar. Sci. 25:101-111.
Marshall, T. C. 1965. Fishes of the Great Barrier Reef and coastal
waters of Queensland. Livingston Publ. Co.. Narberth, PA.
Martin, F. D., AND G. E. Drewry. 1978. Development of the fishes
of the Mid-Atlantic Bight. An atlas of egg, larval and juvenile stages.
Vol. VI. Stromateidae through Ogcocephalidae. U.S. Fish Wildl.
Serv. Biol. Serv. Prog. FWS/OBS-78/12.
Martin, R. A., .AND J. H. FiNHCANE. 1969. Reproduction and ecology
ofthelongnose killifish. Q. J. Fla. Acad. Sci. 31(2):101-l 1 1.
Martinez, A. M. 1983. Identification of brook, brown, rainbow, and
LITERATURE CITED
719
cutthroat trout larvae. Unpubl. MS Thesis. Colorado State Univ.,
Ft. Collins, CO.
Maruyama, K. 1970. Some deep-water fishes from off the Tohoku
and adjacent regions. Bull. Tohoku Reg. Fish. Res. Lab. 30:43-66.
(In Japanese.)
Marx, H., AND D. B. Rabb. 1972. Phyletic analysis of fifty characters
of advanced snakes. Fieldiana Zool. 63:1-321.
Masterman, a. T. 1901. A contribution to the life histories of the
cod and whiting. Trans. R. Soc. Edinb. 40:1-14.
Masuda, H., C. Araga and T. Yoshino. 1975. Coastal fishes of
Southern Japan. Tokai Univ. Press, Tokyo.
Masuda, S., and T. Ozawa. 1979. Reexamination of the holotypes
of Bregmaceros japonicus Tanaka and B. neclabamis Whitley. Jpn.
J. Ichthyol. 25:266-268.
Masuda, T., and K. Tanaka. 1962. Young of labroid and scaroid
fishes from the central Pacific of Japan. J. Tokyo Univ. Fish. 48:
1-98.
Masurekar, v. B. 1968. Eggs and development stages of Tylosurus
crocodilus (LesueuT). J. Mar. Biol. Assoc. India. 9:70-76.
Matarese, A. C. AND S. F. Borton. MS. Reproductive biology and
development of the Pacific spiny lumpsucker, Eumicrolremus orbis
(Cyclopteridae).
, AND J. B. Marliave. 1982. Larval development of laboratory-
reared rosylip sculpin, Ascelichthvs rhodonis (Cottidae). Fish. Bull.
U.S. 80:345-355.
, S. L. Richardson AND J. R. Dunn. 1981. Larval development
of Pacific tomcod, Microgadus pro.ximiis. in the northeast Pacific
Ocean with comparative notes on larvae of walleye pollock, Ther-
agra chakogramma. and Pacific cod, Gadus macrocephalus (Gad-
idae). Fish. Bull. U.S. 78:923-940.
, AND B. Vinter. MS. Development and occurrence of larvae
of the longfin Irish lord, Hemilepidotus zapus (Cottidae).
Mather, F. J. 1958. A preliminary review of the amberjacks, genus
Senota. of the western Atlantic. Proc. 3rd Int. Game Fish Conf
Matias, J. R. 1982. Embryonic diapause in annual fishes: Evaporative
water loss and survival. Experientia (Basel) 38:1315-1316.
Matsubara, K. 1936. A review oftwo genera of the Japanese sculpins,
Ereiinias and Marukawichthys with special reference to their sys-
tematic positions. J. Imp. Fish. Inst. (Tokyo) 31:97-1 14.
. 1942. On the metamorphosis of a clupeoid fish Plerothnssus
gissu Hilgendorf J. Imp. Fish. Inst. (Tokyo) 35(I):1-16.
. 1943. Studies on the scorpaenoid fishes of Japan. Trans. Sig-
enkagaku Kenyusho. Nos. 1 and 2.
. 1955. Fish morphology and hierarchy. Part II. Ishizaki-Sho-
ten, Tokyo.
, AND T. IwAi. 1952. Studies on some Japanese fishes of the
family Gempylidae. Pac. Sci. 6:193-212.
, AND . 1958. Anatomy and relationships of the Japanese
fishes of the family Gempylidae. Mem. Coll. Agnc. Kyoto Uni\
Fish Ser. Spec. No., p. 23-54.
, AND . 1959. Description of a new sandfish, Kraemana
sexradiata. from Japan, with special reference to its osteology. J.
Wash. Acad. Sci. 49:27-32.
, AND A. OcHiAi. 1955. A revision of the Japanese fishes of the
family Platycephalidae (the flatheads). Mem. Coll. Agric. Kyoto
Univ. 68:1-109.
, , K. Am,aok.a and I. Nak,amura. 1964. Revisional
study of the trachinoid fishes of the family Champsodontidae from
the waters around Japan, and Tonking Bay. Bull. Misaki Mar. Biol.
Inst. Kyoto Univ. 6:1-20.
Matsui, K. 1949. Illustration of the normal courses of development
in the fish, Oryzias latipes. Jikken-keitai Gaku 5:33-42. (In Jap-
anese.)
Matsumoto, W. M. 1968. Morphology and distnbution of larval
wahoo Acanthocybium solandn (Cuvier) in the central Pacific Ocean
fishery. U.S. Fish Wildl. Serv. Fish. Bull. 66:299-322.
, E. H. Ahlstrom, S. Jones, W. L. Klawe, W. J. Rr hards and
S. Ueyanagi. 1972. On the clarification of larval tuna identifi-
cation particularly in the genus Thunnus. Fish. Bull. U.S. 70:1-12.
Matsuoka, M., and T. Iwai. 1983. Adipose fin cartilage found in
some teleostean fishes. Jpn. J. Ichthyol. 30:37-46.
Matsuura, K. 1979. Phytogeny ofthesuperfamilyBalistoidea (Pisces:
Tetraodontiformes). Mem. Fac. Fish. Hokkaido Univ. 26:49-169.
. 1982. A list ofthe fishes collected in the Palau and Yap Islands.
Proc. Jpn. Soc. Syst. Zool. 23:80-89.
Matsuura, Y. 1975. A study of the life history of Brazilian sardine,
Sardinetla brasiliensis. III. Development of sardine larvae. Bol.
Inst. Oceanogr. 23:17-29.
, AND M. Katsur.aoawa. 1981. Larvae and juveniles of grey
triggerfish, Balisies capriscus. from Southern Brazil. Jpn. J. Ich-
thyol. 28:267-275.
Mattson.C. R., AND B. L. Wing. 1978. Ichthyoplankton composition
and planktonic volumes from inland coastal waters of southeastern
Alaska, April-November 1972. NOAA Tech. Rep. NMFS SSRF
723.
Maul, G. E. 1951. Monografia dos Peixes do Museu Municipal do
Funchal, Familia Macrouridae e Merlucciidae. Bol. Mus. Munic.
Funchal No. 5 (Art. 12):5-55.
. 1976. The fishes taken in bottom trawls by R.V. Meteor during
the 1967 Seamount Cruises in the Northeast Atlantic. Meteor For-
schungsergeb. 22:1-69.
, AND E. KoEFOED. 1950. On a new genus and species of mac-
rourid fish, Phalacromacrurus panthennus. Ann. Mag. Nat. Hist.,
Ser. 12, 3:970-976.
May, R. C. 1970. Feeding larval marine fishes in the laboratory: a
review. Calif Coop. Oceanic Fish. Invest. Rep. 14:76-83.
. 1974. Larval mortality in marine fishes and the critical period
concept, p. 3-19. In: The eariy life history offish. J. H. S. Blaxter
(ed.). Springer Verlag, Berlin.
, D. Popper and J. P. McVey. 1974. Rearing and larval de-
velopment oi Siganus canaliculatus (Park) (Pisces: Siganidae). Mi-
cronesica 10:285-298.
Mayer, F. 1933. Cuhanichthys cubensis Eigenmann. Blatt. Aquar.-
Terranenk. 44(22):375-376.
Mayer, G. F. 1972. Systematics, functional anatomy, and ecology of
the cardinalfish genus Epigonus (Apogonidae). Unpubl. PhD dis-
sertation. Harvard Univ., Cambridge, MA.
Mayo, C. A. 1973. Rearing, growth, and development of the eggs and
larvae of seven scombrid fishes from the straits of Florida. Unpubl.
PhD thesis. University of Miami, Coral Gables, FL.
Mair, E. 1969. Principles of systematic zoology. McGraw-Hill Publ.,
New York.
. 1981. Biological classification: toward a synthesis of opposing
methodologies. Science 214:510-516.
Mazza, F. 1902. La Lebias calarilana Bon. Segmenlazione dell'uovo
e stadi successivi di sviluppo. Atti Soc. Ligust. Sci. Nat. Genova
13:181-223.
M.AZZARELLi, G. 1917. Contributo alia conosccnza dcllc mctamorfosi
deW Eretmophonis kleinenbergi Gigl. Pubbl. Staz. Zool. Napoli 2.
McAllister, D. E. 1963. A revision of the smelt family, Osmeridae.
Natl. Mus. Can. Bull. 191.
. 1968. Evolution of branchiostegals and classification of tel-
eostome fishes. Natl. Mus. Canada Bull. 221.
, AND J. E. Randall. 1975. A new species of centrolophid fish
from Easter Island and Rapa Iti Island in the South Pacific. Natl.
Mus. Nat. Sci. (Ottawa) Publ. Biol. Oceanogr. 8.
McCann, C, AND D. G. M( Knight. 1980. The marine fauna of New
Zealand: macrourid fishes (Pisces: Gadida). N.Z. Oceanogr. Inst.
Mem. 61:1-91.
McCosKER, J. E. 1977. The osteology, classification and relationships
ofthe eel family Ophichthidae. Proc. Calif. Acad. Sci. 4 1( 1 ): 1-1 23.
. 1979. The snake eels (Pisces. Ophichthidae) of the Hawaiian
Islands, with the descnption of two new species. Proc. Calif. Acad.
Sci. 42(2):57-67.
1982. A new genus and two new species of remarkable Pacific
worm eels (Ophichthidae, subfamily Myrophinae). Proc. Calif Acad.
Sci. 43:59-66.
— , AND R. H. Rosenblatt. 1972. Eastern Pacific snake-eels of
720
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
the genus Callechelys (Apodes: Ophichthidae). Trans. San Diego
Soc. Nat. Hist. 17:15-24.
McCuLLOCH, A. R. 1926. Report on some fishes obtained by the F.I. S.
"Endeavor" on the coasts of Queensland. New South Wales. Vic-
toria. Tasmania, South and south-western Australia. Part V, p. 157-
216. Commonw. Aust. Dep. of Trade and Customs.
McCuLLV, H. 1961. The comparative anatomy of the scales of the
serranid fishes. Unpubl. PhD dissertation. Stanford Univ., Stan-
ford. CA.
. 1970. Amputation and replacement of marginal spines in
ctenoid percoid scales. Proc. Calif Acad. Sci. 38:41 1-414.
McDowALL, R. M. 1968a. The ?>\a\\x^ of Nesogalaxias neocaledonicus
(Weber and de Beaufort) (Pisces: Gala.xiidae). Breviora 286:1-8.
. 1 968b. Gataxias maculatus (Jenyns). the New Zealand white-
bait. Fish. Bull. N.Z. 2:1-84.
. 1 969. Relationships of galaxioid fishes with further discussion
of salmoniform classification. Copeia 1969:796-824.
. 1970. The galaxiid fishes of New Zealand. Bull. Mus. Comp.
Zool. Harv. Univ. 139:341-431.
. 1971a. Fishes of the family Aplochitonidae. J. R. Soc. N.Z.
1:37-52.
. 1971b. The galaxiid fishes of South America. Zool. J. Linn.
Soc. 50:33-73.
. 1973a. The status of the South African galaxiid (Pisces: Gal-
axiidae). Ann. Cape Prov. Mus. Nat. Hist. 9(5):91-101.
. 1973b. Limitation of the genus Sraf/!i'ga/aA7a.s Eigenmann,
1928 (Pisces: Galaxiidae). J, R. Soc. N.Z. 3:193-197.
. 1973c. Relationships and taxonomy of the New Zealand tor-
rent fish, Chehnarrichlhys foslen Haast (Pisces: Mugiloididae). J.
R. Soc. N.Z. 3:199-217.
— . 1976. Fishes of the family Prototroctidae (Salmoniformes).
Aust. J. Mar. Freshwater Res. 27:641-659.
— . 1978. New Zealand freshwater fishes— a guide and natural
history. Heinemann Educational Books, Auckland.
— . 1979. Fishes of the family Retropinnidae (Pisces: Salmoni-
formes)—a taxonomic revision and synopsis. J. Soc. N.Z. 9:85-
121.
— . 1980. Freshwater fishes and plate tectonics in the southwest
Pacific. Palaeogeogr. Palaeoclimatol. Palaeoecol. 31:337-351.
— . 1982. The centrolophid fishes of New Zealand (Pisces: Stro-
mateoidei). J. R. Soc. N.Z. 12:103-142.
— , AND G. A. Eldon. 1 980. The ecology of whitebait migrations.
Fish. Bull. N.Z. 20:1-171.
— , andR. S. Frankenberg. 1981. The galaxiid fishes of Australia
(Pisces: Galaxiidae). Rec. Aust. Mus. 33:433-605.
-, D. A. Robertson AND R. Saito. 1975. Occurrence of galaxiid
larvae and juveniles in the sea. N.Z.J. Mar. Freshwater Res. 9:1-9.
McDowell, S. B. 1973. Order Heteromi (Notacanthiformes), p. 1-
228. In: Fishes of the western North Atlantic. D. M. Cohen (ed.).
Mem. Sears Found. Mar. Res. No. 1 (Pt. 6).
McGiNNis, R. F. 1982. Biogeography of lantemfishes (Myctophidae)
south of SO-S. Biol. Antarct. Seas 12, Antarct. Res. Ser. 35:1-1 10.
McIntosh, W. C. 1892. Contributions to the life-histones and de-
velopment of the food- and other fishes. 10th Annu. Rep. Fish.
Board Scotl., Pt. 3:273-322.
. 1893. Contributions to the life-histones and development of
the food- and other fishes. Scotl. Fish. Board Annu. Rep. 1 1(1892):
239-249.
, AND A. T. Mastermann. 1897. The life-histories of the British
marine food fishes. Cambridge Univ. Press, London.
, AND E. E. Prince. 1890. On the development and life-histories
of the teleostean food- and other fishes. Trans. R. Soc. Edinb. 35:
665-946.
McKenney, T. W. 1959. A contribution to the life history of the
squirrel fish Holocenlrus vexillarius Poey. Bull. Mar. Sci. Gulf Car-
ibb. 9:174-221.
. 1961. Larval and adult stages of the stromateoid fish Pie/?«
regulus with comments on its classification. Bull. Mar. Sci. Gulf
Caribb. 11:210-236.
, E. C. Alexander AND G. S. Voss. 1958. Early development
and larval distribution of the carangid fish, Caranx crysos (Mitchill).
Bull. Mar. Sci. Gulf Caribb. 8:167-200.
Ml Kenzie, R. A., AND S. N. TiBBO. 1963. An occurrence of the opah,
Lampris reg;i«(Bonnaterre) in the northwest Atlantic. J. Fish. Res.
Board Can. 20:1097-1099.
M< Kern, J. L., H. F. Morton AND K. V. KosKi. 1974. Development
of steelhead trout (Salmo gairdneri) otoliths and their use for age
analysis and for separating summer from winter races from hatchery
stocks. J. Fish. Res. Board Can. 31:1420-1426.
M( Lellan, T. 1977. Feeding strategies of the macrourids. Deep-Sea
Res. 24:1019-1036.
Mc Mahon, T. E., AND J. C. Tash. 1979. Effects of formalin (buffered
and unbuffered) and hydrocholoric acid on fish otoliths. Copeia
1979:155-156.
McMillan, C. W. 1979. The embryological development of .Rn'M/wi
marmoratus (Pisces: Cyprinodontidae). Unpubl. MS thesis. College
of Charleston.
McMillan, H. M. 1961. An addition to the knowledge of the fish
Retropinna anisodon Stokell (Retropinnidae). Trans. R. Soc. N.Z.
Zool. 1(10):139-144.
Mead, G. W. 1957. An Atlantic record of the zeoid fish Parazen
pacificus. Copeia 1957:235-237.
. 1963. Observations on fishes caught over the anoxic waters
of the Cariaco Trench, Venezuela. Deep-Sea Res. 10:251-257.
. 1972. Bramidae. Dana-Rep. Carlsberg Found. 81.
, E. Bertelsen AND D. M. Cohen. 1964. Reproduction among
deep sea fishes. Deep-Sea Res. 1 1:569-596.
Meek, S. E., AND S. F. Hildebrand. 1923. The marine fishes of Pan-
ama. Field Mus. Nat. Hist. Publ. Zool. Ser. 15(215).
Meldrim, J. W. 1980. Novumbra hubbsi. p. 128. In: Atlas of North
American freshwater fishes. D. S. Lee, C. R. Gilbert, C. H. Hocutt,
R. E. Jenkins, D. E. McAllister and J. R. Staufler, Jr. (eds). North
Carolina State Mus. Nat. Hist., Raleigh, NC.
Menasse, V, 1970. The spined loach, Cobilis taenia. Trop. Fish Hob-
by. 1 8(6):36-37, 40-41,44.
Menon, A. G. K. 1977. A systematic monograph of the tongue soles
of the genus Cynoglossus Hamilton-Buchanan (Pisces: Cynoglos-
sidae). Smithson. Contrib. Zool. 238.
Mercado S., J. E., and A. Ciardelli. 1972. Contribucion a la mor-
fologia y organogenesis de los leptocefalos del sabalo Megalops
allanlicus (Pisces: Megalopidae). Bull. Mar. Sci. 22:153-184.
Merrett, N. R. 1963. Pelagic gadoid fish in the Antarctic. Nor. Hval-
fangst-tidsskr. 52:245-247.
. 1978. On the identity and pelagic occurrence of larval and
juvenile stages of rattail fishes (family Macroundae) from 60°N,
20°W and 53°N, 20°W. Deep-Sea Res. 25:147-160.
. 1980. Balhytyphlops sewe/i {Pisces: Chlorophthalmidae) a se-
nior synonym of B. azorensis, from the eastern North Atlantic with
notes on its biology. Zool. J. Linn. Soc. 68:99-109.
, J. Badcock and p. J. Herring. 1973. The status of Ben-
ihatbetla infans (Pisces. Myctophoidei), its development, biolu-
minescence, general biology and distribution in the eastern North
Atlantic. J. Zool. Lond. 170:1-48.
, AND N. B. Marshall. 1981. Observations on the ecology of
deep-sea bottom-living fishes collected off northwest Africa (08°-
27°N). Prog. Oceanogr. 9:185-244.
-, Y. I. Sazonov and Y. N. Shcherbachev. 1983. A new genus
and species of rattail fish (Macrouridae) from the eastern North
Atlantic and eastern Indian Ocean, with notes on its ecology. J.
Fish Biol. 22:549-561.
Meshkov, M. M. 1967. The phases in the development of Lola lata
(Linne). Izv. Vses. Nauchno-Issled. Inst. Ozem. Rechn. Rybn. Khoz.
62:181-194.
Methot, R. D. 1981. Spatialcovariationofdailygrowlh rates of larval
northern anchovy, Engraulis mordax. and northern lampfish,
Stenobrachnis leucopsarus. Rapp. P. V. Reun. Cons. Int. Explor.
Mer 178:424-431.
LITERATURE CITED
721
, AND D. Kramer. 1979. Growth of nonhem anchovy, En-
graulis mordax. larvae in the sea. Fish. Bull. U. S. 77:1 124-1 127.
Mettee, M. F., Jr.. and E. C. Beckman. HI. 1978. Notes on the
breeding behavior, embryology and larval development of Cyprin-
odon vanegatus Lacepede in aquana. Tulane Stud. Zool. Bot. 20(3-
4):137-I48.
Mever-Rochow. V. B. 1974. Leptocephali and other transparent fish
larvae from the south-eastern Atlantic Ocean. Zool. Anz. 192(3/
4):240-25l.
MicHENER, C. 1977. Discordant evolution and the classification of
allodapine bees. Syst. Zool. 26:32-56.
Mi< KEVK H, M. F. 1982. Transformation series analysis. Syst. Zool.
31:461-478.
Middauoh, D. P., G. I. Scott AND J. M. Dean. 1981. Reproductive
behavior of the Atlantic silverside, Menidia mcnidia (Pisces, Ath-
erinidac). Environ. Biol. Fishes 6:269-276.
Miller, D. 1958. A key to some of the more common larval fishes
of the Gulf of Maine. NMFS Woods Hole Lab. Misc. Rep. 58- 1 .
, AND R. R. Marak. 1959. The early larval stages of the red
hake, Urophycis chuss. Copeia 1959:248-250.
. AND R. R. Marak. 1962. Early larval stages of the fourspot
flounder. Paralichthys oblongus. Copeia 1962:454-455.
Miller, D. J. 1952. Development through the prelarval stage of ar-
tificially fertilized eggs of the Pacific sardine (Sardinops caeiidca).
Calif. Fish Game 38:587-595,
, AND R. N. Lea. 1972. Guide to the coastal marine fishes of
California. Calif Dep. Fish Game Fish Bull. 157.
Miller, G. L., and S. C. Jorgensen. 1973. Meristic characters of
some marine fishes in the western Atlantic Ocean. Fish. Bull. U.S.
71:301-312.
Miller. J. M., and B. Y. SuMiDA. 1974. Development and life history
of Caranx mate (Carangidae). Fish. Bull. U.S. 72:497-514.
, AND J. W. Tucker. Jr. 1979. X-radiography of larval and
juvenile fishes. Copeia 1797:535-538.
. W. Watson and J. M. Leis. 1979. An atlas of nearshore
marine fish larvae of the Hawaiian Islands. Sea Grant Misc. Rep.
UNlHl-SEAGRANT-MR-80-02. Univ. Hawaii Sea Grant College
Program. Honolulu.
Miller. P. J. 1961. Evidence for the breeding in Manx waters of the
sand-smelt. Athenna preshvler C & V. Rep. Mar. Biol. Sta. Point
Erin 74:27-28.
. 1973. The osteology and adaptive features of Rhyacichlhys
aspro (Teleostei: Gobioidei) and the classification of gobioid fishes.
J. Zool. (Lond.) 171:397-434.
Miller, R. R. I960. Systematics and biology of the gizzard shad
(Dorosoma cepedianum) and related fishes. U.S. Fish Wildl. Serv.
Fish. Bull. 60:371-392.
, AND J. C. Briggs. 1962. Daclylos^opus ainnis, a new sand
stargazer from rivers of the Pacific slope of southern Mexico. Occas.
Pap. Univ. Mich. Mus. Zool. 627.
Miller, R. V. 1969. Constancy of epibranchial organs and fourth
epibranchial bones within species groups of clupeid fishes. Copeia
1969:308-312.
, AND J. W. VAN Landingham. 1969. Additional procedures
for effective enzyme clearing and staining of fishes. Copeia 1969:
829-830.
MiLLiKEN, D. M. 1975. Species of the genus Bregmaceros in the Gulf
of Mexico and Caribbean Sea, including the description of a new
species from the Cariaco Trench. Unpubl. MS thesis. Univ. South
Florida, Tampa, FL.
, and E. D. Holde. 1984. A new species of Bregmacerotidae
(Pisces), Bregmaceros canton, from the western Atlantic Ocean.
Bull. Mar. Sci. In press.
MiiwARD, N. H. 1966. Development of eggs and early larvae of the
Australian smelt Rctropinna semoni (Weber). Proc. Linn. Soc.
N.S.W. 90:218-227.
MiNAMi, T. 1981a. The early life history ofa flounder I'seudorbombus
pentophlhalmus. Bull. Jap. Soc. Sci. Fish. 47:849-856.
. 1981b. The early life history ofa sole Heteromycteris japon-
icus. Bull. Jap. Soc. Sci. Fish. 47:857-862. (In Japanese. English
abstract.)
. 1981c. The early life history ofa flounder Limanda yoko-
hamae. Bull. Jpn. Soc. Sci. Fish. 47:141 1-1419.
. 1982a. The early life history ofa flounder Pleuronkhthys
cornutus. Bull. Jap. Soc. Sci. Fish. 48:369-374. (In Japanese. En-
glish abstract.)
. 1 982b. The early life history of a tongue fish Paraplagusia
japonica. Bull. Jap. Soc. Sci. Fish. 48:1041-1046.
MiNOPRio. J. L. 1944. Considerations sobre el desarrollo embrionano
del Perjerrey. Acta. Zool. Lilloana 2:9-14.
MisiTANO, D. 1978. Description of laboratory-reared larvae of the
Buffalo Sculpin. Enophrvs bison (Pisces, Cottidae). Copeia 1978:
635-642.
. 1980. A record of internal fertilization in the roughback scul-
pin Chitonotus pugetensis with descriptive notes of their early lar-
vae. Copeia 1980:162-165.
MiTANi, F. 1960. Fishery biology of the yellowtail, Seriola qumque-
radiata T. & S., inhabiting the water around Japan. Mem. Fac.
Agric. Kinki Univ. 1:81-300.
Mitchell, C P.. AND B. P. Penlington. 1982. Spawning of Ga/a.v(aj
fascialus Gray (Salmoniformes: Galaxiidae). N.Z. J. Mar. Fresh-
water Res. 16:131-133.
, AND D. ScoTT. 1979. Muscle myogens in New Zealand Gal-
axiidae. N.Z. J. Mar. Freshwater Res. 13:285-294.
MiTO, S. 1954. Breeding habits ofa blennioid fish, Salarias enosimae.
Jpn. J. Ichthyol. 3:144-152, (In Japanese. English summary.)
. 1955. Breeding habits of a percoid fish. P/M/op5Sf»;e(o/;. Sci.
Bull. Fac. Agric. Kyushu Univ. 15(l):95-99.
. 1956a. On the egg development and hatched larvae of Lelh-
rinus nematacanthus Bleeker. Sci. Bull. Fac. Agric. Kyushu Univ.
15:497-500.
. 1956b. On the egg development and hatched larvae of Op/e^-
nathus fascialus (Temmni:}/. et Schlegel). Sci. Bull. Fac. Agric. Kyu-
shu LJniv. 15:501-506.
. 1957a. Egg development and hatched larvae of G/>W/a pi/nc-
tata Gray (Girellidae, Teleostei). Jpn. J. Ichthyol. 6:105-108. (In
Japanese, English summary.)
. 1957b. On the egg development and larvae of a Japanese sea
bass, Lateolabrax japonicus (Cuvier). Sci. Bull. Fac. Agric. Kyushu
Univ. 16:1 15-124. (In Japanese. English summary.)
. 1958. Studies on the eggs, larvae and juvenile of Japanese
fishes. Series I. In: K. Uchida et al. 1958. Fish Dep.. Fac. Agric.
Kyushu Univ.. Fukuoka. Japan. (In Japanese.)
. 1960. Egg development and hatched lar\ae of the common
dolphin-fish. Coryphaena hippunis Linne. Bull. Jap. Soc. Sci. Fish.
26(3):223-226. (In Japanese. English abstract.)
. 1961a. Pelagic fish eggs from Japanese waters— I. Clupeina.
Chanina, Stomiatina, Myctophida, Anguillida, Belonida and Syng-
nathida. Sci. Bull. Fac. Agric. Kyushu Univ. 18:285-310.
. 1961b. Pelagic eggs from Japanese waters— II. Lamprida, Zei-
da, Mugilina, Scombrina, Carangina and Stromateina. Sci. Bull.
Fac. Agric. Kyushu Univ. 18:451-465.
— . 1962a. Pelagic fish eggs from Japanese waters— V. Calliony-
mina and Ophidiina. Sci. Bull. Fac. Agric. Kyushu Univ. 19:377-
380.
. 1962b. Pelagic fish eggs from Japanese waters— VI. Labrina.
Sci. Bull. Fac. Agric. Kyushu Univ. 19:493-502. (In Japanese. En-
glish summary.)
— . 1962c. Pelagic fish eggs from Japanese waters— VII. Chaeto-
donlina. Balistina and Ostraciontina. Sci. Bull. Fac. Agric. Kyushu
Univ. 19:503-506. (In Japanese, English summary.)
— . I962d. On the egg development and early lanae ofa trachi-
noid fish, Champsodon snyden (Franz). Bull. Jpn. Soc. Sci. Fish.
28(5):499-503.
— . 1963. Pelagic fish eggs from Japanese waters— III. Percina.
VIII. Cottina. IX. Echeneida and PIcuroneclida. Jpn. J. Ichthyol.
11:39-102.
— . 1966. Fish eggs and larvae. In: Illustrated encyclopedia of the
722
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
marine plankton of Japan, Vol. 7, S. Motoda (ed.). Soyo-Sha, To-
kyo. (In Japanese.)
. 1979a. A perspective on morphology and phylogeny offish
eggs and larvae. Kaiyo Kaguku (Mar. Sci.) 1 1:87-92. (In Japanese.)
. 1979b, Fish eggs, Kaiyo Kaguku (Mar. Sci.) 1 1:126-130. (In
Japanese.)
, AND K. UcHiDA. 1958. On the egg development and hatched
larvae of a scorpaenoid fish, Pterois lunulata Temminck et Schlegel.
Sci. Bull. Fac. Agric. Kyushu Univ. 16:381-385. (In Japanese, En-
glish summary,)
, M. Ukawa AND M. HioucHi. 1967. On the larval and young
stages of a serranid fish, Epmephelus akaara (Temmmck et Schle-
gel). Bull. Naikai Reg. Fish. Res, Lab. 25:337-347.
MocHioKA,N.,S. K^KLiDA andO. Tabeta. 1982. Congrid leptocepha-
li in the western North and Middle Pacific. I. Exterilium Ariosoma-
type larvae, J, Fac. Appl. Biol. Sci, Hiroshima Univ, 21:35-66,
Moffatt, N, M. 1981. Survival and growth of northern anchovy
larvae on low zooplankton densities as affected by the presence of
a Chlorella bloom. Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:
475-480.
MoHR, E. 1926. Die Gattung Ze«aR'/iop/en« Gill. Zool. Jahrb. Jena
Abt. Syst, 52:231-266,
, 1936a. Hemirhamphiden-Studien IV,-VI, IV. Die Gattung
Dermogenys van Hasselt. Mitt. Zool. Mus. Bed. 21(l):34-55.
. 1936b. Hemirhamphiden-Studien IV.-VI. V. Die Gattung
Nomorhainphus Weber & de Beaufort. Mitt. Zool. Mus. Berl. 21(1):
55-58.
. 1936c. Hemirhamphiden-Studien IV.-VI. VI. Die Gattung
Hemirhamphodon Bleeker 1866. Mitt. Zool. Mus. Berl. 21(1):59-
64.
1937. Revision der Centnscidae (Acanthopterygii Centrisci-
formes) (in German). Dana-Rep. Carlsberg Found. 18.
MoHSiN, A, K. M. 1978. The osteology of the grey mullet, /.;ia ii/^-
vindis (Valenciennes). The branchiocranium, vertebral column, and
girdles. Malayan Nat. J. 31(4):203-218.
MoK, H. K., AND S. C. Shen. 1983. Osteology and phylogeny of
Squamipinnes. Spec. Publ. Taiwan Mus. Zool. Ser. I,
MoNOD, T. 1968. Le complexe urophore des poissons teleosteens.
Mem. Inst. Fondam. Afr. Noire. 81.
MONTALENTI, G. 1933. Famiglia 5: Maenidae, p. 383-384. In: Uova,
larve e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli
Monogr. 38.
. 1937a. Famiglia 6: Mullidae, p, 391-398. In: Uova, larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
1937b. Cepolidae, p. 407-4 1 0. In: Uova, larve e stadi giovanili
di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
MoNTOLio, M. A, 1976, Estudio taxonomico y morfometrico de los
estadios larvales de dos especies de Carangidae Decapterus punc-
tatus (Agassiz, 1829) y Caranx crysos (Mitchill, 1815) y su distri-
bucion en el Golfo de Mexico. Rev. Invest. Inst. Nac. Pesca 2:85-
125.
MooK, D. 1977. Larval and osteological development of the sheeps-
head, Archosargitsprobalocephalus {Pisces: Sparidae). Copeia 1977:
126-133.
MooKERjEE, H. K. 1946. Identification of eggs of common freshwater
fishes of Bengal. Sci. Cuh. 1 1(1):48.
, AND S, R, Mazumdar, 1950. Some aspects of the life history
oi Clarias batrachus (Linn.). Proc. Zool. Soc. Bengal 3(l):71-83.
Moore, C. J. 1980. Spawning of Menidia menidia (Pisces: Atherini-
dae). Copeia 1980:886-887.
Moore, D, 1962, Development, distribution, and comparison of rud-
der fishes Kyphosus sectatnx (Linnaeus) and A', incisor (Cuvier) in
the western North Atlantic, U, S, Fish Wildl. Serv. Fish. Bull. 196:
451-480,
MoREiRA, C. 1920. Recherches sur la reproduction de l'//op//ai ma/-
abaricus (Bloch) et sur I'lncubation d'oeufs de Salmo fario au Bresil.
Bull. Soc. Zool. Fr. 44:329-336,
MoRELAND, J. 1960. A new species and genus of congiopodid fish
from southern New Zealand. Rec. Dom. Mus. (Wellington) 3:241-
246,
Moreno, C, A. 1980. Observations on food and reproduction in Tre-
matomus bernacchii (Pisces: Nototheniidae) from the Palmer Ar-
chipelago, Antarctica. Copeia 1980:171-173.
Mori, K, 1984, Early life history of Luljanus villa (Lutjanidae) in
Yuya Bay, the Sea of Japan. Jpn. J. Ichthyol. 30:374-392.
, S. Uevanagi AND Y. NisHiKAWA. 1971. The development of
artificially fertilized and reared larvae of the yellowfin tuna, Thun-
nus albacares. Bull. Far Seas Fish. Res. Lab. (Shimizu) 5:219-232.
MoRiN, R. P., andK. W. Able. 1983. Patterns of geographic variation
in the egg morphology of the fundulid fish, Fundulus hcteroditus.
Copeia 1983:726-740,
MoRiTA, S. 1984. History of the hemng fishery and review of the
artificial propagation technique for herring in Japan. Can. J. Spec.
Publ. Fish. Aquat. Sci. In press.
MoRK, J., P. SoLEMDAL AND G. Sl'ndness. 1983. Identification of
marine fish eggs: a biochemical genetics approach. Can. J. Fish.
Aquat. Sci. 40:361-369.
MoRovic, D. 1953. Sur la determination des mugesadnatiquesd'apres
la forme de I'otolith sagitta. Notes Inst. Oceanogr. Split 9:1-7.
Morris, D, E,, and R, K^navama. 1964-1969. Life history study of
the Moi, Polydactylus sextiUs. Job Completion Rep., Proj. No.
F-S-R-1 1. Div. Fish Game, Honolulu, Hawaii. Mimeogr.
Morris, R, 1951. Early development of the cottid fish Clinocottus
recalvus (Greeley). Calif Fish Game 37:281-300.
. MS. Some notes on the early life history of the night smelt,
Spirinchus slarksi (Fisk).
Morrow, J. E., Jr. 1964a. Family Chauliodontidae, p. 274-289. /«.
Fishes of the western North Atlantic. H. B. Bigelow (ed.). Mem.
Sears Found. Mar. Res. No. 1 (Pt. 4).
. 1964b. Stomiatidae, p. 290-310. In: Fishes of the western
North Atlantic. H. B. Bigelow (ed). Mem. Sears Found. Mar. Res.
No. 1 (Pt. 4).
. 1964c. Family Malacosteidae, p, 523-549. In: Fishes of the
western North Atlantic. H. B. Bigelow (ed). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 4).
. 1980. The freshwater fishes of Alaska. Alaska Northwest Publ.
Co., Anchorage,
, and R. H. Gibbs, Jr.
1964, Family Melanostomiatidae, p.
351-510. In: Fishes of the western North Atlantic. H. B. Bigelow
(ed.). Mem. Sears Found. Mar. Res. No. 1 (Pt. 4). (Genera Metan-
oslomias and Eustomias by R. H. Gibbs, Jr,, the remainder by J.
E. Morrow, Jr.)
, AND S. J. Harbo. 1 969. A revision of the sailfish genus Islioph-
orus. Copeia 1969:34-44.
MosER, H. G. 1972. Development and geographical distnbution of
the rockfish, Sebasles macdonaldi (Eigenmann and Beeson, 1893),
family Scorpaenidae, off southern California and Baja California.
Fish. Bull. U.S. 70:941-958.
. 1974. Development and distribution of larvae and juveniles
oi Sebaslolobus {Pisces; Family Scorpaenidae). Fish. Bull. U.S. 72:
865-884.
. 1981. Morphological and functional aspects of marine fish
larvae, p. 89-131. In: Marine fish larvae. Morphology, ecology and
relation to fisheries. R. Lasker (ed). Univ. Washington Press, Se-
attle.
, and E. H, Ahlstrom. 1970. Development of lantemfishes
(family Myctophidae) in the California Current. Part I. Species with
narrow-eyed larvae. Nat. Hist. Mus. Los Ang. Cty. Sci. Bull. 7.
, and . 1972. Development of the lantemfish, Scopelop-
sis multipunclalus Brauer 1 906, with a discussion of its phylogenetic
position in the family Myctophidae and its role in a proposed
mechanism for the evolution of photophore patterns in lantem-
fishes. Fish. Bull. U.S. 70:541-564,
, AND . 1974. Role of larval stages in systematic inves-
tigations of marine teleosts: the Myctophidae, a case study. Fish,
Bull, U,S. 72:391-413.
LITERATURE CITED
723
.AND . 1978. Larvae and pelagic juveniles of blackgill
rockfish. Sehastes melanoslomus. taken in midwaler trawls off
southern California. J. Fish. Res. Board Can. 35:981-996.
, . D. Kramer and E. G. Stevens. 1973. Distribution
and abundance of fish eggs and larvae in the Gulf of California.
Cal. Coop. Oceanic Fish. Invest. Rep. 17:1 12-128.
, andE. M.Sandknop. 1977. Guide to the identification
of scorpionfish larvae in the eastern Pacific with comparative notes
on species of Sebastes and Helicolemis from other oceans. NOAA
Tech. Rep. NMFS Circ. 402.
— . AND J. L. Butler. 1981. Description ofreared larvae and early-
juveniles of the calico rockfish, Sebastes dallii. Calif. Coop. Oceanic
Fish. Invest. Rep. 22:88-95.
, AND . MS. Descriptions of reared larvae of six species
of Sebastes. In: Proc. of a workshop on widow rockfish {Sebastes
entomelas). W. H. Lenarz and D. Gunderson (eds.).
MosHER, C. 1954. Observations on the spawning behavior and the
early larval development of the Sargassum fish, Histno histno (Lin-
naeus). Zoologica (N.Y.) 39:141-152.
MoiiLTciN, L. 1970. The 1970 longfin smelt population spawning run
in Lake Washington with notes on egg development and changes
in the population since 1964. LInpubl. MS thesis. Univ. Washing-
ton, Seattle.
Mover, J. T. 1979. Mating strategies and reproductive behaviour of
ostraciid fishes at Miyake-jima. Japan. Jpn. J. Ichthyol. 26:148-
160.
MujiB. K. A. 1967. The cranial osteology of the Gadidae. J. Fish. Res.
Board Can. 24:1315-1375.
. 1969. Cranial osteology of Brosme bromse with notes on its
position within the subfamily Lotinae (Gadidae). J. Fish. Res. Board
Can. ItAlX-AJ,!.
Mi'KHACHEVA, V. A. 1957. Notes on the development of the Far
Eastern na\aga(Eleginus gracilis1'\\e^\\K). Tr. Inst. Okeanol. Akad.
Nauk SSSR 20:356-370. (In Russian. Transl. by Am. Inst. Biol.
Sci., 1959, p. 291-302.)
. 1964. The composition of species of the genus Cyclothone
(Pisces, Gonoslomatidae) in the Pacific Ocean, p. 98-146. In: Fish-
es of the Pacific and Indian Oceans. Biology and Distribution. T.
S. Rass(ed.). Tr. Inst. Okean. 73. (English translation 1966 by Isr.
Program Sci. Transl.)
. 1978. A review of the species of Diplophos Giinther (Gon-
ostomatidae, Osteichthyes) and their vertical and geographical dis-
tribution. Trans. P. P. Shirshov Inst. Ocean. 1 1 1:10-27. (In Rus-
sian, English abstract.)
, AND O. A. ZviAGiNA. 1960. Development of the Pacific cod,
Gadus morhua macrocephalus Tilesius. Tr. Inst. Okeanol. Akad.
Nauk SSSR 31:145-165. (Transl. from Russian by W. E. Ricker,
1969, Fish. Res. Board Can.. Transl. Ser. 393.)
MOller, H., AND G. Sterba. 1963. Elektronenmikroskopische Un-
tersuchungen iiber Bildung und Struktur der EihiiUen bei Knock-
enfischen. II. Die Eihiillen jungerer und alterer Oozytcn von Cy-
no/eft(a5 /)W/<:)K;VSteindachner(Cyprinodontidae). Zool. Jahrb. Abt.
Anal. 80(4):469-488.
Munk, O. 1966. Ocular anatomy of some deep-sea teleosts. Dana-
Rep. Carlsberg Found. 70.
. AND E. Bertelsen. 1980. On the esca light organ and its
associated light-guiding structures in the deep-sea anglerfish Chaen-
ophryne draco. Vindensk. Medd. Dan. Naturhist. Foren. 142:103-
129.
MuNRO. I. S. R. 1945. Postlarval stages of Australian fishes— No. 1.
Mem. Queensl. Mas. 12:148-152.
. 1950. Revision of Bregmaceros with descriptions of larval
stages from Australasia. Proc. R. Soc. Queensl. 61:37-53.
1954. Eggs and larvae of the four-winged flying fish, l/irun-
Dasson i^earf; (Whitley) (Blenniidae). Aust. J. Mar. Freshwater Res.
6:30-34.
. 1967. The Fishes of New Guinea. New Guinea Dep. Agric.
Stock Fish.. Port Moresby.
Mlntz, W. R. a. 1976. On yellow lenses in mesopelagic animals. J.
Mar. Biol. Assoc. U.K. 56:963-976.
Murray, J., and J. Hjort. 1912. The depths of the ocean. Macmillan
and Co., London.
MusicK, J. A. 1973. A meristic and morphometric comparison of the
hakes, Vrophycis chuss, and V. tenuis (Pisces, Gadidae). Fish. Bull.
U.S. 71:479-488.
Musienko, L. N. 1963. Ichthyoplankton of the Bering Sea (data of
the Bering Sea Expedition of 1958-1959), p. 251-286. In: P. A.
Moiseev (ed.). Soviet Fisheries Investigations in the Northeast Pa-
cific, I . Tr. Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz. Okean-
ogs. 48.
Musiv, Y. I., and V. A. Sergi-ienko. 1977. Fingerlings of the genus
Lutjanus (Lutjanidae, Perciformes) from the Gulf of Aden. J. Ich-
thyol. (Engl. Transl. Vopr. Ikhtiol.) 17:151-154.
MiERS, G. S. 1937. The deep-sea Zeomorph fishes of the family Gram-
micolepidae. Proc. U.S. Natl. Mus. 84:145-156.
. 1958. Trends in the evolution of teleostean fishes. Stanford
Ichthyol. Bull. 7(3):27-30.
. 1960. A new zeomorph fish of the Family Oreosomatidae
from the coast of California, with notes on the family. Stanford
Ichthyol. Bull. 7(4):89-98.
, AND W. C. Freihofer. 1966. Megalomyctendae. a previously
unrecognized family of deep-sea cetomimiform fishes based on two
new genera from the North Atlantic. Stanford Ichthyol. Bull.
8:193-206.
, AND C. B. Wade. 1946. New fishes of the families Dactyloscop-
idae, Microdesmidae, and Antennariidae from the west coast of
Mexico and the Galapagos Islands with a brief account of the use
of rotenone fish poisons in ichthyological collecting. Allan Hancock
Found. Pac. Exped. 9(6):I51-179.
Nafpaktitis, B. G. 1975. Review ofthelantemfish genus A'oroscope/iw
(family Myctophidae) in the north Atlantic and Mediterranean.
Bull. Mar. Sci. 25:75-87.
. 1977. Family Neoscopelidae, p. 1-12. /w.Fishesofthe western
North Atlantic. R. H. Gibbs, Jr. (ed). Mem. Sears Found. Mar.
Res. No. l(Pt. 7).
. 1978. Systematics and distribution of lantemfishes of the gen-
dichthys speculiger (Valenciennes). .'\ust. J. Mar. Freshwater Res.
5(l):64-69.
. 1955. Eggs and larvae of the sabre-toothed oyster blenny
era Lobianchia and Diaphiis (Myctophidae) in the Indian Ocean.
Nat. Hist. Mus. Los Ang. Cty. Sci. Bull. 30:1-92.
, and M. Nafpaktitis. 1969. Lantemfishes (family Myctophi-
dae) collected during cruises 3 and 6 of the R'V "Anton Bruun"
in the Indian Ocean. Nat. Hist. Mus. Los Ang. Cty. Sci. Bull. 5:1-
79.
, and J. R. Paxton. 1968. Review of the lantemfish genus
Lampadena with a description of a new species. Contrib. Sci. (Los
Ang.) 138.
, and . 1978. W(o/vc/i/iJ«. a new genus of Myctophidae
based on Diaphus urolampus. Copeia 1978:492-497.
, R. H. Bakus, J. E. Craddcxk, R. L. Haedrich. B. H. Robison
AND C. Karnella. 1977. Family Myctophidae, p. 13-265. In:
Fishes of the western North Atlantic. R. H. Gibbs, Jr. (ed.). Mem.
Sears Found. Mar. Res. No. l(Pt. 7).
Nagiec, C. 1 979. On the developmental stages of the fins and axial
skeleton of whitefish Coregunus lavaretus L. Pol. Arch. Hydrobiol.
26:529-543.
Nair, G. S. 1958. On the breeding habits and development of .4m-
bassis gymnocephalus {Lac). Bull. Cent. Res. Inst. Univ. Travan-
core, India 5C(l):69-76.
N^iR, R. V. 1952a. Studies on some fish eggs and larvae of the Madras
plankton. Proc. Indian Acad. Sci. Sec. B 35:181-208.
. 1952b. Studies on some post-larval fishes of the Madras plank-
ton. Proc. Indian Acad. Sci. Sect. B 35:225-244.
. 1960. Notes on the spawning habits and early life-history of
724
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
the oil sardine, Sardmella longiceps Cuv. & Val. Indian J. Fish.
6:342-359.
Nakabo, T. 1982a. Revision of the family Draconeltidae. Jpn. J.
Ichthyol. 28:335-367.
. 1982b. Revision of genera of the dragonets (Pisces: Callionym-
idae). Publ. Seto Mar. Biol. Lab. 27:77-131.
Nakah,\ra, K. 1962. A note on the larvae of .-l/i/;.?o/»u fu/iro.v Lowe
and Erythrocles schlegeli (Richardson) collected off the southern
coast of Kyushu. Bull. Jpn. Soc. Sci. Fish. 28:484-488.
Nakamlira, L 1974. Some aspects of the systematics and distribution
of billfishes. NOAA Tech. Rep. NMFS SSRF-675(2):45-53.
. 1977. Gempylidae. //;. FAO species identification sheets for
fishery purposes. Western Central Atlantic, Fishing Area 31. Vol.
2. FAO. Rome.
. 1980. New record of a rare gempylid, Thyrsiloides inarleyi.
from the Sea of Japan. Jpn. J. Ichthyol. 26:357-360.
, AND E. Fuji. 1983. A new genus and species of Gempylidae
(Pisces: Perciformes) from Tonga Ridge. Publ. Seto Mar. Biol. Lab.
28(4/6):173-191.
, T. IwAi AND K. Matshbara. 1968. A review of the sailfish.
spearfish. marlin and swordfish of the world. Kyoto Univ. Spec.
Rep. 4.
Nakamlira, M. 1951. Thelifehistory ofacyprinid fish. Opsariichthys
uncirostris (Temminck et Schlegel) in Lake Biwa. Misc. Rep. Res.
Inst. Nat. Resour. (Tokyo) 19-21:70-78.
Nakamura, M. N. 1969. Cyprinid fishes of Japan: Studies on the life
history of cyprinid fishes of Japan. Res. Inst. Nat. Resour. (Tokyo)
Spec. Publ. 4.
Nakamura, S. 1936. Larvae and young of fishes found in the vicinity
of Kominato, II-IV. II. Athenon elynnis loTdiin and Starks. J. Imp.
Fish. Inst. (Tokyo) 30:219-225.
. 1937. Larvae and young of fishes found in the vicinity of
Kominato, VII-VIII. J. Imp. Fish. Inst. (Tokyo) 32:17-25.
Naplin, N. A., andC. L. Obenchain. 1980. A description of eggs and
larvae of the snake eel, Pisodonophis crueiilifer (Ophichlhidae).
Bull. Mar. Sci. 30:413-423.
Nash, C. E. 1973. Automated mass-production of Anemia salina
nauplii for hatcheries. Aquaculture 2:289-298.
, C.-M. Kiio, W. D. MA15I5EN AND C. L. Paulsen. 1977. Swim
bladder inflation and survival of Mugil cephatus to 50 days. Aqua-
culture 12:89-94.
Nassogne, A. 1970. Influence of food organisms on the development
and culture of pelagic copepods. Helgol. Wiss. Meeresunters. 20:
333-345.
Nellen, W. 1973a. Untersuchungen zur Verteilung von Fischlarven
und Plankton im Gebiet der Grossen Meteorbank. Meteor For-
schungsergeb. 13:47-69.
. 1973b. Fischlarven des Indischen Ozeans. Meteor For-
schungsergeb. Biol. 14:1-66. (English summary.)
, AND G. Hempel. 1970. Beobachtungen am Ichthyoneuston
der Nordsee. Ber. Dtsch. Wiss. Komm. Meeresforsch. 21:31 1-348.
Nellis, D. W. 1980. Reproduction in the ocean triggerfish Cantht-
dennis sufflamen. Caribb. J. Sci. 16(1-4): 1 67.
Nelsen.O. E. 1953. Comparative embryology of the vertebrates. Blak-
iston, New York.
Nelson, G. J. 1 966a. Osteology and relationships of the eel, Neenche-
lys huilendijki. Copeia 1966:321-324.
. 1966b. Gill arches of leleostean fishes of the order Anguilli-
formes. Pac. Sci. 20:391-408.
. 1967. Gill arches of teleostean fishes of the family Clupcidac.
Copeia 1967:389-399.
. 1969. Gill arches and the phylogeny of fishes, with notes on
the classification of vertebrates. Bull. Amer. Mus. Nat. Hist. 141:
479-552.
. 1970. The hyobranchial apparatus of teleostean fishes of the
families Engraulidae and Chirocentridae. Am. Mus. Novit. 2410.
. 1972. Cephalic sensory canals, pitlines, and the classification
of esocoid fishes, with notes on galaxiids and other teleosts. Am.
Mus. Novit. 2492:1-49.
— . 1973. Relationships of clupeomorphs, with remarks on the
structure of the lower jaw in fishes, p. 333-349. In: Interrelation-
ships of fishes. P. H. Greenwood. R. S. Miles and C. Patterson
(eds.). Zool. J. Linn. Soc. 53 (Suppl. 1).
. 1978. Ontogeny, phylogeny, palaeontology, and the biogenetic
law. Syst. Zool. 27:324-345.
. 1983. Anchoa argouixitlala with notes on other eastern Pacific
anchovies and the Indo-Pacific genus Encrasicholina. Copeia 1983:
48-54.
. 1984. Identity of the anchovy Hildebrandichlhys setiger with
notes on relationships and biogeography of the genera Engraulis
and Cetengraulis. Copeia 1984:422^27.
, AND N. Platnick. 1981. Systematics and biogeography. Cla-
distics and vicariance. Columbia Univ. Press, New York.
Nelson. J. S. 1975. Records of a new form of the marine cottid fish
Anltpodocoltus gatatheac from the east coast of New Zealand. Natl.
Mus. N.Z. Rec. l(4):80-86.
. 1976. Fishes of the world. John Wiley & Sons. New York.
. 1 977. Fishes of the Southern Hemisphere genus Neophrynich-
ihys (Scorpaeniformes: Cottoidei), with descriptions of two new
species from New Zealand and Macquarie Island. J. R. Soc. N.Z.
7:485-511.
. 1978. Limnichlhys polyaclis. a new species of blennioid fish
from New Zealand, with notes on the taxonomy and distribution
of other Creediidae (including Limnichthyidae). N.Z. J. Zool.
5:351-364.
. 1982. Two new South Pacific fishes of the genus Ehinania
and contributions to the systematics of Psychrolutidae (Scorpaen-
iformes). Can. J. Zool. 60:1470-1504.
Nichols, J. H. 1971. Pleuronectidae. Cons. Int. Explor. Mer. Fiches
d'identification des oeufs et larves de poissons. No. 4-6:2-18.
Nichols, J. T., and C. M. Breder. Jr. 1927. The marine fishes of
New York and southern New England. Zoologica (N.Y.) 9(1):1-
192.
, AND C. M. Breder. Jr. 1928. An annotated list of the Synen-
tognathi with remarks on their development and relationships. Col-
lected by the Arcturus. Zoologica (N.Y.) 8:423-448.
Nielsen, J. G. 1963a. On the development of Cheilodaclylus varie-
.?<3«« Valenciennes 1 883 (Cheilodactylidae). Copeia 1963:528-533.
. 1963b. Notes on some Heterosomata (Pisces) from NW South
America with the description of a new genus and species and a new
subspecies of Paralichthinae. Vidensk. Medd. Dan. Naturhist. For-
en. 125:375-400.
. 1963c. Description of two large unmetamorphosed flatfish
larvae (Heterosomata). Vidensk. Medd. Dan. Naturhist. Foren 1 25:
401-406.
. 1964. Fishes from depths exceeding 6000 meters. Galathea
Rep. 7:113-124.
. 1966. Synopsis of the Ipnopidae (Pisces, Iniomi). Galathea
Rep. 8:49-75.
. 1969. Systematics and biology of the Aphyonidae (Pisces,
Ophidioidea). Galathea Rep. 10:7-88.
. 1973. Stylephondae. p. 335. //I. Check-list of the fishes of the
north-eastern Atlantic and of the Mediterranean (Clofnam). Vol. I.
J. C. Hureau and T. Monod (eds.). UNESCO, Pans.
, .'VND J.-C. HuREAi'. 1980. Revision of the ophidiid genus
Spectninculus. a senior synonym of Parabassogigas (Pisces, Ophi-
diiformes). Steenstrupia 6(1 1): 149-1 69.
, AND D. G.Smith. 1978. The eel family Nemichthyidae(Pisces,
Anguilliformes). Dana-Rep. Carlsberg Found. 88:1-71.
NiKOLSKY, G. V. 1961. Special ichthyology. Second revised and en-
larged edition (translated from Russian, original date of publication
1954). Jerusalem. (Not seen; cited by Tyler, 1968.)
. 1963. The ecology of fishes. Academic Press. London.
NisHiKAWA, Y. 1979. Early development of the double-lined mackerel,
Grammatorcymis bicarinatiis (Quoy and Gaimard), from the west-
em tropical Pacific. Bull. Far Seas Fish. Res. Lab. (Shimizu) 17:
125-140.
. 1982. Early development of the fishes of the family Gempy-
LITERATURE CITED
725
lidae. I. Larvae and juveniles of escolar, Lepidocybium flavobnin-
neum (Smith). Bull. Far Seas Fish. Res. Lab. (Shimizu) ( 19): 1-13.
, AND L N.AKAMLiRA. 1978. Post larvae and juveniles of the
gempylid fish Neoepinnuta onentalis (Gilchrist and Von Bonde),
from the North Arabian Sea. Bull. Far Seas Fish. Res. Lab. (Shim-
izu) (16):75-91.
NiSHiMURA. S. 1964. Additional mformation on the biology of the
dealfish, Trachiplerus ishikawai Jordan and Snyder. Bull. Jpn. Sea
Reg. Fish. Res. Lab. 13:127-129.
. 1966. Early life history of the deep-sea smelt, Glossarwdon
semifasciatus (Kishinouye) (Teleostei: Clupeida). Pt. 1. Publ. Seto
Mar Biol. Ub. 13:349-360.
NoAKES, D. L. G., AND E. K. Balon. 1982. Life histories of tilapia:
an evolutionary perspective, p. 61-68. In: biology and culture of
tilapia. R. S. V. Pullin and R. H. Lowe-McConnell (eds.). ICLARM
Conference Proceedings 7. International Center for Living Aquatic
Research Management, Manila, Philippines.
NoRCROSS, J. J., W. H. Massmann and E. B. Joseph. 1961. Investi-
gations of inner continental shelf waters off lower Chesapeake Bay.
Part 2. Sand lance larvae, Ammodvles amencanus. Chesapeake Sci.
2:46-59.
. S. L. Richardson, W. H. Massman and E. B. Joseph. 1974.
Development of young bluefin (Pomalonms saltalnx) and distri-
bution of eggs and young in Virginia coastal waters. Trans. Am.
Fish. Soc. 103:477-497.
Norden, C. R. 1961. Comparative osteology of representative sal-
monid fishes with particular reference to the grayling {Thymallus
arcticus) and its phylogeny. J. Fish. Res. Board Can. 18:679-791.
Norman, J. R. 1926a. A report on the flatfishes (Heterosomata) col-
lected by the F. I. S. "Endeavour," with a synopsis of the flatfishes
of Australia and a revision of the Rhombosoleinae. Biol. Res. Fish-
ing Exped. F. I. S. Endeavour 5:217-308.
. 1926b. The development of the chondrocranium of the eel
(Anguilta vulgaris) with observations on the comparative mor-
phology and development of the chrondrocranium in bony fishes.
Philos. Trans. R. Soc. Lond. B. Biol. Sci. 214:369-464.
. 1929. The teleostean fishes of the family Chiasmodontidae.
Ann. Mag. Nat. Hist. 3(10): 529-544.
. 1930. Oceanic fishes and flatfishes collected in 1925-1927.
Part I. Oceanic fishes. Part II. Flatfishes. Discovery Rep. 2:261-
370.
. 1934. A systematic monograph of the flatfishes (Heteroso-
mata). Vol. I . Pseltodidae, Bothidae. Pleuronectidae. Br. Mus. (Nat.
Hist.) Lond.
. 1937. Coast fishes, part 2, the Patagonian Region. Discovery
Rep. 16:1-50.
. 1938a. Coast fishes, part 3, the Antarctic Zone. Discovery
Rep. 18:1-104.
. 1938b. On the aflinities of the Chilean fish, Normanichthvs
crocken C\avk. Copeia 1938:29-32.
. 1966. A draft synopsis of the orders, families and genera of
recent fishes and fish-like vertebrates. Br. Mus. (Nat. Hist.) Lond.
North, A. W., AND M. G. White. 1982. Key to fish postlarvae from
the Scotia Sea, Antarctica. Cybium, 3rd Ser. 6(1): 13-32.
Nibelin,0. 1947. Antarctic fishes. Sci. Results Norw. Antarct. Exped.
26.
. 1948. Fishes collected by the "Skagerak" Expedition in the
eastern Atlantic 1946. Goteb. Vetensk. Samh. Handl. 6:1-95.
. 1951. Subantarctic and Antarctic fishes. Sci. Results Brategg
Exped. 2.
Obrhelova.N. 1961. Vergleichende Osteologie der tertiaren Susswas-
serfische Bohmens (Gobioidei). Sb. Ustred, LIstavu Geol. 26:103-
192.
OcHiAi. A., AND K. Amaora. 1963. Description of larvae and young
of four species of flatfishes referable to subfamily Bothidae. Bull.
Jpn. Soc. Sci. Fish. 29(2): 127-1 34. (In Japanese.)
, AND Y. Nozawa. 1980. On the metamorphosis and burrowing
habits of the snake eel, Muraenichthvs gymnolus. Jpn. J. Ichthyol.
27:237-242.
O'CoNNELL. C. P. 1953. The life history of the cabezon, Scorpaenich-
thys mannoralus (Ayres). Calif Dep. Fish Game Fish Bull. 93:1-
10.
. 1976. Histological criteria for diagnosing the starving con-
dition in early post yolk-sac larvae of the northern anchovy En-
graulis mordax Girard. J. Exp. Mar. Biol. Ecol. 25:285-312.
. 1980. Percentage of starving northern anchovy, Engraiilis
mordax. larvae in the sea as estimated by histological methods.
Fish. Bull. U.S. 78:475-489.
. 1981a. Development oforgan systems in the northern anchovy
Engraulis mordax and other teleosts. Am. Zool. 21:429-446.
. 1981b. Estimation by histological methods of the percent of
starving northern anchovy larvae in the sea. Rapp. P. V. Reun.
Cons. Int. Explor. Mer 178:357-360.
O'Day, W. T. 1972. The histology and fine structure of some biolu-
minescent organs in deep sea fishes. Unpubl. PhD dissertation.
Univ. South. Calif, Los Angeles.
. 1974. Bacterial luminescence in the deep-sea anglerfish OnW-
rodes acanthias {GWhen. 1915). Contnb. Sci. (Los Ang.) 255:1-12.
Oelschlager, H. a. 1 974. Das Caudalskelett von Lampns regius und
seine Ableitung von I'elUer hvpselopterus. Senckenb. Biol. 55( 1/3):
77-85.
. 1976a. On the evolution and ecological adaptations of the
AUotriognathi. Rev. Trav. Int. Peches Mant. 40(3, 4):69 1-694.
. 1976b. Gotteslachse und Bandfische — Erforschungsgeschichle
und Evolution der Glanzfischartigcn. Nat. Mus. l06(2):48-54.
. 1978a. Comparative morphological and functional investi-
gations on the pelvis of the oarfishes (Teleostei: AUotriognathi).
Zool. Beitr. 24(2): 165-192.
. 1978b. On the morphology and biomechanics of .4g/'o«rt'/!?/!V'.s
Phillipps in comparison with other AUotriognathi (Pisces, Teleos-
tei). Zool. Jahrb. Abt. Anat. 100:118-153.
. 1979. On the functional osteology of the Lophotidae (Pisces:
Teleostei). Zool. Jahrb. Abt. Anat. 102:336-360.
Ogasawara. Y.. S. Akatsu AND Y, Taki. 1978. Juvenile stages and
effect of salinity on the survival of larvae and juveniles in the striped
fingerfish, Monodaclylus sebae. Jpn. J. Ichthyol. 24:246-250.
Ohsima, Y., AND N. Nakamiira. 1941. Note on the life history of
Triacanthus breviroslns Temminck et Schlegel. Bull. Jpn. Soc. Sci.
Fish. 10(4): 1 7 1-1 76. (In Japanese. English summary.)
OiESTAD, v., AND E. MoKSNESs. 1981. Study of growth and survival
of herring larvae (Chipea harengiis) using plastic bag and concrete
basin enclosures. Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:
144-149.
OiokDA, Y. 1960. Studies on the freshwater fishes of Japan. II. Special
part. J. Fac. Fish. Prefect. Univ. Mie 4(2):267-588.
, AND R. Semshi. 1938. Studies on the early life history of 9
species of freshwater fishes of Japan. Bull. Biogeogr. Soc. Jpn. 8(15):
223-263.
Okamura, O. 1970a. Studies on the macrouroid fishes of Japan —
morphology, ecology and phylogeny. Rep. Usa Mar. Biol. Stn. 17(1-
2):1-I79.
. 1970b. Fauna Japonica: Macrourina (Pisces). Academic Press
Japan.
Okivama, M. 1963. Larvae and young of the witch flounder, Glyp-
locephalus stelleri (Schmidt) at metamorphosis stages. Bull. Jpn.
Sea Reg. Fish. Res. Lab. 1 1:101-108.
. 1964. Early life history of the Japanese blanquillos, Brart-
chioslegus japoniciis japonicus (Hounuyn). Bull. Jap. Sea Reg. Fish.
Res. Lab. 13:1-14. (In Japanese, English abstract.)
. 1967. Study on the early life history of a flounder. Paralichthys
olivaceus (Temminck et Schlegel). I. Descnptions of postlarvae.
Bull. Jpn. Sea Reg. Fish. Res. Lab. 17:1-12.
. 1970. Studies on the early life history of the rainbow runner,
Elagalis bipinnulatus (Quoy and Gaimard) in the Indo-Pacific
Oceans. Bull. Far Seas Fish. Res. Lab. (Shimizu) 3:167-186.
. 1971. Early life history of the gonostomatid fish. Maurolicus
muellen (Gmelin), in the Japan Sea. Bull. Jpn. Sea Reg. Fish. Res.
Lab 23:21-53.
726
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
. 1972. Morphology and identification of the young ipnopid.
"Macristiella" from the tropical western Pacific. Jpn. J. Ichthyol.
19:145-153.
— . 1974a. Studieson the early life history of a flounder, Para/;W;-
ihys olivaceus (Temminck et Schlegel). II. Descriptions of juveniles
and the comparison with those of the related species. Bull. Jpn. Sea
Reg. Fish. Res. Lab 25:39-61.
. 1974b. The larval taxonomy of the primitive myctophiform
fishes, p. 609-621. In: The early life history offish. J. H. S. Blaxter
(ed.). Springer- Verlag, New York.
. 1979a. Manuals for the larval fish taxonomy. ( 1 ) Definition
and classification of larval stages. Kaiyo to Seibutsu (Ocean and Its
Life) 1:54-59. (In Japanese.)
. 1979b. Larval morphology and interrelationships of the myc-
tophoid fishes. Kaiyo Kaguku (Mar. Sci.) 11:131-145. (In Jap-
anese.)
. 1980. Manuals for the larval fish taxonomy (5). Diversity of
Myctophiformes larvae. Aquabiology 2:124-129.
. 1981. A larval Ipnops and its possible metamorphosing pro-
cess. Jpn. J. Ichthyol. 28:247-253.
— . 1982a. Manuals for the larval fish taxonomy (9). Percoid
larvae and spination. Oceans Mammals 4(2):92-99. (In Japanese,
English abstract.)
. 1982b. Embryoof ,-l//o/ep«/;o//a«(y/ Jordan and Hubbs, 1925.
Aquabiology 4(4):cover.
, AND H. Sando. 1976. Early life history of sea raven. Bull. Jap.
Sea Reg. Fish. Res. Lab. 27:1-10.
, AND K. Takahashi. 1976. Larval stages of the right eye floun-
ders (subfamily Pleuronectinae) occurring in the Japan Sea. Bull.
Jpn. Sea Reg. Fish. Res. Lab. 27:1 1-34.
, AND S. Ueyan.agi. 1977. Larvae and juvenile of the Indo-
Pacific dogtooth tuna, Gymnosarda umcolor (Ruppell). Bull. Far
Seas Fish. Res. Lab. (Shimizu) 15:35-49.
, AND . 1978. Interrelationships of scombroid fishes: an
aspect from larval morphology. Bull. Far Seas Fish. Res. Lab.
(Shimizu) 16:103-113.
Olney, J. E., AND G. C. Grant. 1976. Early planktonic larvae of the
blackcheek tonguefish, Symphurus plagiusa (Pisces: Cynoglossi-
dae), in the lower Chesapeake Bay. Chesapeake Sci. 1 7(4):229-237.
, AND D. F. Markle. 1979. Description and occurrence of
vexillifer larvae of Echiodon (Pisces: Carapidae) in the western
Norih Atlantic and notes on other carapid vexilHfers. Bull. Mar.
Sci. 29:365-379.
, AND A. Naplin. 1 980. Eggs of the scalloped ribbon-fish, Zu
cristatus, (Pisces, Trachipteridae) in the western North Atlantic.
Copeia 1:165-166.
-, .^ND G. R. Sedberrv. In press. Dragonet larvae (Teleostei,
Callionymidae) in continental shelf waters off the eastern United
States. Biol. Oceanogr.
Olsen, v., AND D. Merriman. 1946. Studies on the marine resources
of southern New England. IV. The biology and economic impor-
tance of the ocean pout, Macrozoarces amencanus (Bloch and
Schneider). Bull. Bingham Oceanogr. Collect. Yale LJniv. 9(4): 1-
184.
Orcutt, H.G. 1950. The life history of the starry flounder, P/ar/f/!//;,r5
stellatus (Pallas). Calif Dep. Fish Game Fish Bull. 78.
Orellana, M. C, AND F. Balbontin. 1983. Estudio comparativo de
las larvas de Clupeiformes de la costa de Chile. Rev. Biol. Mar.
(Chile) 19:1-46.
Orton, G. L. 1953a. Development and migrations of pigment cells
in some teleost fishes. J. Morphol. 93:69-99.
. 1953b. The svstematics of vertebrate larvae. Syst. Zool. 2:63-
75.
. 1955a. Separation of eggs of synenthognath and allotriognath
fishes in early embryonic stages. Calif Fish Game 41:103-105.
. 1955b. Early developmental series of the California barracuda,
Sphyraena argentea Girard. Calif Fish Game 41:167-176.
. 1955c. The role of ontogeny in svstematics and evolution.
Evolution 9:75-83.
— . 1955d. Early developmental stages of the California scor-
pionfish, Scorpaena guttata. Copeia 1955:210-214.
— . 1957. The bearing of larval evolution on some problems in
frog classification. Syst. Zool. 6:285-292.
— . 1964. The eggs of scomberesocid fishes. Copeia 1964:144-
150.
O'TociLE. M. J. 1977. Development and distnbutional ecology of the
larvae of the west coast sole Auslroglossus microlepis. Fish. Bull.
S. Afr. 9:32-45.
, AND D. P. F. King. 1974. Early development of the round
herring Etrumeus teres (De Kay) from the southeast Atlantic. Vie
Milieu 24:443-452.
Owen, R. 1866. On the anatomy of the vertebrates. Vol. 1. Fishes
and Reptiles. Longman Green. London.
Owen, R. W. 1980. Patterning of flow and organisms in the larval
anchovy environment. lOC Workshop Rep. No. 28:167-200.
OzAWA, T. 1976. Early life history of the gonostomatid fish, Pollichthys
mauli. in the oceanic region off southern Japan. Jpn. J. Ichthyol.
23:43-54.
. 1978. Notosudid fish larvae in the ocean off southern Japan.
Jpn. J. Ichthyol. 25:79-88.
. 1983. Study on the early life history and the distribution of
oceanic fish larvae in the ocean ofl" southern Japan. Unpubl. PhD
dissertation. Kyushu Univ.
, AND S. Matslh, 1979. First record of the schindlerid fish,
Schmdleria praematura. from southern Japan and the south China
Sea. Jpn. J. Ichthyol. 25:283-285.
Padmanabhan, K. G. 1961. The early development of Sotenostomus
cyanopterus. Bull. Cent. Res. Inst. Trivandrum 80:1-13.
Padoa, E. 1947. Forme post-larvali e giovanili di Carapus. Pubbl.
Staz, Zool. Napoli 20:1 1 1-120.
. 1956a. Famiglia 4: Trichiuridae. p. 508-513. In: Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
. 1956b. Famiglia 10-13: Stromateidae, Centrolophidae. No-
meidae, Tetragonuridae, p. 534-547. In: Uova, larvc, e stadi giov-
anili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1956c. Divisione: Carangiformes, p. 548-572. /«.■ Uova larve
e stadi giovanili di Telostei. Fauna Flora Golfo Napoli Monogr.
38.
. 1956d. Sottordine: Chromides, p. 573-576. In: Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
. 1956e. Triglidae, Peristediidae, Dactylopteridae, Gobiidae,
Echneidae, Jugulares, Gobiesocidae, Heterosomata, Pediculati. p.
627-888. In: Uova. larve, e stadi giovanili di Teleostei. Fauna Flora
Golfo Napoli Monogr. 38.
. 1956f Famiglia: Gobiidae, p. 648-678. In: Uova, larva e
stadi giovanili de Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. I956g. Famiglia 1: Trachinidae, p. 687-697. //;. Uova. larvae
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
. 1956h. Famiglia 6: Clinidae, p. 745-752. In: Uova, larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1956i. Famiglia 8: Ophidiidae, p. 755-761. In: Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
. 1956j. Famiglia 9: Carapidae. p. 761-774. In: Uova, larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1956k. Heterosomata, p. 783-877. In: Uova, larve e stadi
giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
Pal, R. N., and H. A. Khan. 1969. Breeding and development of
Heteropneustes fossdis (Bloch). Proc. Indian Sci. Congr. 56:543-
544.
Palko, B. J., G. L. Beardsley and W. J. Richards. 1981. Synopsis
of the biology of the swordfish, Xiphias gladius Linnaeus. NOAA
Tech. Rep. NMFS Circ. 441.
, AND W. J. Richards. 1969. The rearing of cowfish. Salt water
Aquar. Mag. 5(3):67-70.
LITERATURE CITED
727
Palmer. G. 1961. The dealfishes (Trachipteridae) of the Mediterra-
nean and northeast Atlantic. Bull. Br. Mus. (Nat. Hist.) Zool.
7:335-351.
. 1973. Lamprididae. Regalecidae, Trachiptendae, Radii-
cephalidae, Lophotidae, p. 328-334. In. S. C. HureauandT. Monod
(eds.). Check-list of the fishes of the north-eastern Atlantic and of
the Mediterranean (Clofnam). Vol. I. UNESCO. Paris.
Palomera. I., AND J. M. FoRTLiNO. 1981. Larvas Icptocefalas dc peces
anguilliformes en la costa noroccidental de Africa. Res. Exped.
Cient. 9:3-15.
. AND P. Rubies. 1977. Descnpcion de huevos y larvas de
Microchirus ocellatus y M. aze\ia (Pleuronectiformes. Soleidae) de
las costas del NW de Africa. Res. Exped. Cient. B/O Comide
6:211-220.
Panchen, a. L. 1982. The use of parsimony in testing phylogenetic
hypotheses. Zool. J. Linn. Soc. 74:305-328.
Pannella, G. 1971. Fish otoliths: daily growth layers and penodical
patterns. Science (Wash. D.C.) 173:1124-1127.
. 1980. Growth patterns in fish sagittae, p. 519-560. /«. Skeletal
growth of aquatic organisms; biological records of environmental
change. D. C. Rhoads and R. A. Lutz (eds.). Plenum Press, New
York.
Papadopoulu, C, G. D. Kanias and E. Moraitopoulou-Kassimati.
1978. Zinc content in otoliths of mackerel from the Aegean. Mar.
Pollut. Bull. 9:106-108.
, and . 1980. Trace element content in fish oto-
liths in relation to age and size. Mar. Pollut. Bull. 1 1:68-72.
Parenti. L. R. 1981. A phylogenetic and biogeographic analysis of
cyprinodontiform fishes (Teleostei, Atherinomorpha). Bull. Am.
Mus. Nat. Hist. 168:335-557.
. 1984. On the relationships of phallostethid fishes (Atherino-
morpha), with notes on the anatomy of Phalloslethus dunckeri Re-
gan, 1913. Am. Mus. Novit. 2779.
Parin, N. V. 1961. The basis of classification of the flying fishes
(families Oxyporhamphidae, Exocoetidae). Tr. Inst. Okeanol. Akad.
NaukSSSR 43:92-183.
. 1967. Review of the marine Belonidae of the western Pacific
and Indian oceans. Tr. Inst. Okean. Akad. Nauk SSSR 84:3-83.
(In Russian: Natl. Mar. Fish. Serv. Syst. Lab. Transl. No. 68.)
. 1968. Scomberesocidae (Pisces, Synentognathi) of the eastern
Atlantic Ocean. Atl. Rep. 10:275-290.
. 1976. Two new species of Benthodesmus (Trichiuridae, Os-
teichthyes) from the western Tropical Pacific Ocean. Tr. Inst.
Okeanol. Akad. Nauk SSSR 104:191-194. (In Russian, English
abstract.)
. 1977. New species of A/e/aA!Oi?;^wa (Zoarcidae, Osteichthy-
es), M. inexpeclatum, from the western equatorial Pacific. Tr. Inst.
Okeanol. Acad. Nauk SSSR 107:63-67. ^
. 1978. New records of midwater fishes from off New Guinea
and Tonga Islands with descriptions of two new species of Eusto-
mias (Family Melanostomiatidae) and Benthodesmus (Family
Trichiundae). Tr. Inst. Okeanol. Akad. Nauk SSSR 111:156-168.
(In Russian, English Abstract.)
, and D. A. AsTAKHov. 1982. Studies on the acoustico-lateralis
system of beloniform fishes in connection with their systematics.
Copeia 1982:276-291.
, and V. E. Bekker. 1972. Materials on taxonomy and distn-
bution of some trichiuroid fishes (Pisces, Trichiuroidae: Scombro-
labracidae, Gempylidae. Trichiuridae). Proc. P. P. Shirshov Inst.
Oceanol. 93:1 10-204 (In Russian, English abstract.)
, and T. N. Belvanina. 1972. New data on the distribution
and morphology offish of the genus Macnstwlla. J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 12:1018-1022.
, B. B. CoLLETTE and Yu. N. Shcherbachev. 1 980. Prehminary
review of the marine halfbeaks (Hemiramphidac, Beloniformes) of
the tropical Indo-West Pacific. Tr. Inst. Okean. Akad. Nauk SSSR
97:7-173. (In Russian, English abstract.)
, and N. N. Gorbunova. 1964. On the reproduction and de-
velopment of some synentognalhs (Pisces, Beloniformes) from the
Indian Ocean (based on material from the R/S "Vityaz"). Tr. Inst.
Okean. Akad. Nauk SSSR 73:224-234.
, AND Y. I. KuKUiEv. 1983. Re-establishment of the validity
oi Lampns tmmaculata Gilchrist and the geographical distribution
of Lampridae. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 23(1):1-
12.
, AND S. Mekhailin. 1981. A new species of cutlassfish. Le/j;-
dopus dubnis Parin and Mikhailin (Trichiuridae), from the eastern
tropical Atlantic Ocean. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
21(3):l-8.
, and G. N. Pokhil'skaya. 1968. The age variability and dis-
tribution of a rare oceanic fish Eumectchthys fiski (Pisces, Lophoti-
dae). Probl. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 8:808-812.
Parker, G. H. 1903. The optic chiasma in teleosts and its bearing on
the asymmetry of the Heterosomata (flatfishes). Bull. Mus. Comp.
Zool. Harv. 40:221-242.
Parker. K. R. 1980. A direct method for estimating norihem ancho-
vy, Engraulis inordax, spawning biomass. Fish. Bull. U.S. 78:541-
544.
Parr, A. E. 1926. Investigations on the Cyclopterini. Bergens Mus.
Aarbok 1924-25, Naturvid. Raekke (7).
. 1927. The stomiatoid fishes of the suborder Gymnophoto-
dermi (Astronesthidae, Melanostomiatidae, Idiacanthidae) with a
complete review of the species. Bull. Bingham Oceanogr. Collect.
Yale Univ. 3.
. 1933. Deep-sea Berycomorphi and Percomorphi from the
waters around the Bahama and Bermuda Islands. Bull. Bingham
Oceanogr. Collect. Yale Univ. 3(6).
. 1946. The Macroundae of the western North Atlantic and
Central American seas. Bull. Bingham Oceanogr. Collect. Yale Univ.
10 (Art. l):l-99.
Pati, S. 1979. On the maturation and spawning of Chinese pomfret
Pampus chinensis (Euphrasen) from Orissa coast. Indian J. Fish.
26:150-162.
Patterson, C. 1968. The caudal skeleton in mesozoic acanthopterygi-
an fishes. Bull. British Mus. (Nat. Hist.) Geol. 17:5-102.
. 1981a. Agassiz, Darwin, Huxley, and the fossil record of te-
leost fishes. Bull. British Mus. (Nat. Hist.) Geol. 35:213-224.
. 1981b. Significance of fossils in determining evolutionary re-
lationships. Annu. Rev. Ecol. Syst. 12:195-223.
. 1982. Morphological characters and homology. /n. Problems
of Phylogenetic Reconstruction. K. A. Joysey and A. E. Friday
(eds.). Proceedings of a symposium, Cambridge, England, April
1980. Published for the Systematics Association by Academic Press,
New York. Syst. Assoc. Spec. Vol. 21.
, AND D. E. Rosen. 1977. Review of ichthyodectiform and other
Mesozoic teleost fishes and the theory and practice of classifying
fossils. Bull. Am. Mus. Nat. Hist. 158:83-172.
Paulin, C. D. 1983. A revision of the family Moridae (Pisces: Ana-
canthini) within the New Zealand region. Natl. Mus. N.Z. Rec. 2(9):
81-126.
Paxton, J. R. 1967. A distributional analysis for the lantemfishes
(family Myctophidae) of the San Pedro Basin. California. Copeia
1967:422-440.
. 1972. Osteology and relationships of the lantemfishes (family
Myctophidae). Nat. Hist. Mus. Los Ang. Cty. Sci. Bull. 13:1-81.
. 1979. Nominal genera and species of lantemfishes (family
Myctophidae). Contrib. Sci. (Los Ang.) 322.
Pearcy, W. G. 1962. Egg masses and early developmental stages of
the scorpaenid fish, Sebastolobus. J. Fish. Res. Board Can. 19:11 69-
1173.
Pearse, a. G. E. 1968. Histochemistry — theory and applied. Vol. 2.
Little Brown, Boston.
Pearson, J. C. 1938. Life history of the striped bass, or rockfish,
Roccus saxalilis (Walbaum). Bull. Bur. Fish. Wash. 49(28):825-
851.
. 1941. The young of some marine fishes taken in lower Ches-
apeake Bay. Virginia, with special reference to the gray sea trout
728
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Cynoscion regalis (B\och). U.S. Fish Wildl. Serv. Fish. Bull. 50:79-
102.
Peden, a. E. 1970. A new cottid fish, Nautichthys robustus. from
Alaska and British Columbia. Natl. Mus. Nat. Sci. (Ottawa) Publ.
Biol. Oceanogr. 2.
. 1981. On the identity of two cottid subspecies, /cW;/u« /)!» -
chami burchami and I. h. fuscescens. Copeia 1981:482-484.
, AND M. E. Anderson. 1978. A systematic review of the fish
genus Lycodapus (Zoarcidae) with descriptions of two new species.
Can. J. Zool. 56:1925-1961.
, AND . 1981. Lycodapus (Pisces: Zoarcidae) of eastern
Bering Sea and nearby Pacific Ocean, with three new species and
a revised key to the species. Can. J. Zool. 59:667-678.
, AND C. A. CoRBETT. 1973. Commensalism between a liparid
fish, Careproclus sp. and the lithodid box crab, Liphotilhodes for-
aminatus. Can. J. Zool. 51:555-556.
Pellegrin.J. 1907. Liste des poissons recueillis a Madagascar par M.
F. Geay. Description d'une espece nouvelle. Bull. Mus. Paris 1907:
201-206.
. 1914. Sur un atherinide nouveau de Madagascar appartenant
au genre Bedotia. Bull. Soc. Zool. Fr. 39:178-180.
. 1935. Les poissons africains de la famille des cromeriides et
leurs affinites. Arch. Mus., Ser. 6, 12:461-463.
Penaz, M. 1968. Development of the chub, LpWfKcwi fcp/!a/(« (Lin-
naeus 1758) m the posthatching penod. Zool. Listy 1 7(3):269-278.
. 1975. Early development of the graylmg Thymallus ihymallus
(Linnaeus, 1758). Acta. Sci. Nat. Acad. Sci. Bohemoslov. Brno.
9:1-35.
. andJ. GajdCisek. 1979. Early development of bream,. -l/iranNi
brama. from the water reservoir Mostiste, Czechoslovakia. Folia
Zool. 28(4):347-360.
Penrith, M. J. 1965. A new species of flat fish, Mancopsella mitjordi.
from South Africa, with notes on the genus Mancopsella. Ann. S.
Afr. Mus. 48:181-188.
Penrith, M.-L. 1969. The systematics of the fishes of the family
Clinidae in South Africa. Ann. S. Afr. Mus. 55:1-121.
Peppar, J. L. 1965. Some features of the life history of cockscomb
prickleback, Anoplarchus purpurescens. Unpubl. MS thesis. Univ.
British Columbia, Vancouver.
Perlmutter, a. 1939. An ecological survey of young fish and eggs
identified from two-net collections, p. 11-71. In: A biological sur-
vey of the salt waters of Long Island. Suppl. 28th Ann. Rep., 1938,
New York State Cons. Dep. Albany, No. 15, Pt. 2, Sect. 1.
, L. BoGRAD and J. J. Pruginin. 1957. Use of the estuarine
and sea fish of the family Mugilidae (grey mullets) for pond culture
in Israel. Proc. Fish. Counc. Mediterr., FAG 4:289-304.
Pertseva, T. A. 1936. Key to pelagic eggs of fishes of the Barents Sea.
Pischepromizdat, Moscow.
Pertseva-Ostrolimova, T. A. 1954. Materials on the development
of far eastern flounders (Pleuroneclidae). I. Development of the
yellowfin so\e-Linianda aspera Pallas. Tr. Inst. Okeanol. Acad. Nauk.
SSSR 11:221-232.
. 1961. The reproduction and development of far eastern floun-
ders. Tr. Inst. Okeanol. Akad. Nauk. SSSR. (In Russian; transl.
avail. Fish. Res. Bd. Canada, Transl. Serv. No. 856, 1967.)
. 1964. Some morphological characteristics of myctophid larvae
(Myctophidae, Pisces). In: Fishes of the Pacific and Indian Oceans,
biology and distribution. T. S. Rass (ed.). Tr. Inst. Okeanol. Akad.
Nauk SSSR 73:79-97. (In Russian, English transl., U.S. Dep. Com-
merce 65-5120.)
. 1965. Ratfish larvae from the Gulf of Tonkin. Tr. Inst. Okean-
ol. Acad. Nauk. SSSR 80:177-220. (In Russian.)
. 1967. har\a\ ^\a%es oi s,ome Protomyctophum and Eleclrona
species (Pisces, Myctophidae). Tr. Inst. Okeanol. Akad. Nauk SSSR,
84:222-236. (In Russian, English transl. Bur. Commer. Fish.,
Southwest Fish. Ctr., La Jolla.)
. 1971. Unusual flounder larvae (Pleuronectiformes, Pisces). J.
Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 1 1:817-822.
. 1972. The larvae of lantemfish (Myctophidae) collected by
Australian Expeditions on H.M.A.S. Gascoyne and H.M.A.S. Dia-
mantina. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 12:634-643.
. 1974. New data on lanternfish larvae (Myctophidae, Pisces)
with oval eyes from the Indian and Pacific Oceans. Tr. Inst. Okean-
ol. Akad. Nauk SSSR 96:77-142.
. 1975. Larvae of the lanternfish, D;u/)/h<5 a,?a55/r; (Myctophi-
dae, Pisces). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 15:606-610.
. 1977. On the development of some species of the genus Gv"i-
no.sco/)e/i«(Myctophidae). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
17:58-65.
, AND T. S. Rass. 1973. Ichthyoplankton of the southeastern
Pacific Ocean. Trudy Inst. Okeanol. Acad. Nauk SSSR 94:7-70.
Peters, K. M. 1981. Reproductive biology and developmental os-
teology of the Rorida blenny. Chasmodes saburrae (Perciformes:
Blenniidae). Northeast Gulf Sci. 4(2):79-98.
Peters, N. 1963. Embryonale Anpassungen oviparerZahnkarpfenaus
periodisch austrocknenden Gewassem. Int. Rev. Gesamlen Hy-
drobiol. 48(2):257-313.
Petersen, C. G. J. 1904. On the larval and postlarval stages of the
long rough dab and the genus Pleuronectes. Medd. Komm. Ha-
vunders. Ser. Fiskeri 1(1).
. 1905. On the young stages of the genus Zeugopterus. Rep.
Dan. Biol. Stn. 12:25-32.
. 1906. On the larval and post-larval stages of some Pleuro-
nectidae (Pleuronectes. Zeugopterus). Medd. Komm. Havunders.
Ser. Fiskeri 2(1).
. 1909. On the larval and post-larval stages of some Pleuro-
nectidae (Zeugopterus. .irnoglossus. Solea). Medd. Komm. Ha-
vunders., Ser. Fiskeri 3 (1):1-I8.
. 1917. On the development of our common gobies (G"ofei«)
from the egg to the adult stages etc. Rep. Dan. Biol. Stn. 24:5-16.
. 1919. Our gobies (Gobiidae). From the egg to the adult stages.
Rep. Dan. Biol. Stn. 25:45-66.
Peterson, C. L. 1956. Observations on the taxonomy, biology, and
ecology of the engraulid and clupeid fishes in the Gulf of Nicoya,
Costa Rica. Bull. Inter-Am. Trop. Tuna Comm. Bull. 1:139-280.
Phillips, J. B. 1942. Osteology of the sardine (Sardinops caerulea). J.
Morphol. 70:463-500.
Phonlor, G. 1978. Morfologia e biologia dos ovos de Engraulidae do
Sul do Brasil (Osteichthyes, Engraulidae). Unpubl. thesis. Univ.
Sao Paulo, Brazil.
. 1979. Fecundafao artificial e desenvolvimento dos ovos e
larvas de Parana signala em laboratorio (Osteichthyes, Carangi-
dae). Atl. Rio Grand, 3:47-56.
PiCQUELLE.S. J., andR. P. Hewitt. 1983. Northern anchovy spawning
biomass for the 1982-83 California fishing season. Calif Coop.
Oceanic Fish. Invest. Rep. 24.
Pietri, R. 1983. An ultrastructural study of the chorion and its fila-
ments in Oryzias latipes. Unpubl. MS thesis. Rutgers. State Univ.
New Jersey, New Brunswick.
Pietsch, T. W. 1972. A review of the monotypic deepsea anglerfish
family Centrophrynidae: taxonomy, distribution, and osteology.
Copeia 1972:17-47.
. 1974. Osteology and relationships of deep-sea anglerfishes of
the family Oneirodidae with a review of the genus Oneirodes Liit-
ken. Nat. Hist. Mus. Los Ang. Cty. Sci. Bull. 18:1-113.
. 1975. Systematics and distribution of ceratioid anglerfishes
ofthe genus Chaenophryne (iamWy Oneirodidae). Bull. Mus. Comp.
Zool. Harv. Univ. 147(2):75-I00.
. 1976. Dimorphism, parasitism and sex: Reproductive strat-
egies among deepsea ceratioid anglerfishes. Copeia 1976:781-793.
. 1978a. The feeding mechanisms of Slylephorus chordalus
(Teleostei: Lampriformcs): functional and ecological implications.
Copeia 1978:255-262.
— . 1978b. Evolutionary relationships ofthe sea moths (Teleostei:
Pegasidae) with a classification of gasterosteiform families. Copeia
1978:517-529.
LITERATURE CITED
729
. 1979. Systematics and distribution of the ceratioidanglerfishes
of the family Caulophrynidae. Contrib. Sci. (Los Ang.) 310.
. 1981. The osteology and relationships of the anglerfish genus
Telrabrachium with comments on lophiiform classification. U.S.
Fish. Bull. 79:387-419.
. 1984. The genera of frogfishes (family Antennariidae). Copeia
1984:27^4.
, AND D. B. Grobecker. 1980. Parental care as an alternative
reproductive mode in an antennariid anglerfish. Copeia 1980:551-
553.
, AND . In press. Frogfishes of the world: Systematics,
zoogeography and behavioral ecology. Stanford University Press.
Stanford, CA.
PlVNlCKA, K. 1970. Morphological variation in the burbot (Lora /o/a)
and recognition of the subspecies: a review. J. Fish. Res. Board
Can. 27:1757-1765.
Platt, C, and a. N. Popper. 1981. Fine structure and function of
the ear, p. 3-36. In: Hearing and Sound Communication in Fishes.
W. N. Tavolga. A. N. Popper and R. R. Fay (eds.). Springer-Verlag,
New York.
Podosinnikov, a. Y. 1977. Early ontogeny of the "stnped grunt."
Paraprislipoma humile (Pomadasyidae. Pisces). J. Ichlhyol. (Engl.
Transl. Vopr. Ikhtiol.) 17:683-685.
Poll, M. 1953. Poissons. III. Teleosteens Malacopterygiens. Result.
Scient. Exped. Oceanogr. Beige Eaux Cot. Afr. Atlant. Sud. 4(2).
. 1964. Une famille dulcicole nouvelle de poissons Africains:
les Congothrissidae. Acad. R. Sci. Outre-Mer. Mem. N. S. 15:1-
40.
Ponomareva, L. A. 1949. Eggs and fry of fishes of the Kara Sea. Tr.
Uses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz. Okeanogr. 17:
189-205.
Popper, A. N., and S. Coombs. 1980a. Acoustic detection by fish, p.
403-430. In: Environmental Physiology of Fish. M. A. Ali (ed.).
Plenum Publ. Corp., New York.
, and S. Coombs. 1980b. Auditory mechanisms in teleost fishes.
Am. Sci. 68:429-440.
PtippER, D., H. Gordon and G. W. Kjssil. 1973. Fertilization and
hatching of rabbitfish Siganus rivulatus. Aquaculture 2:37-44.
Poss, S. G. 1975. Infraorbital bones in cottid and hexagrammid fishes.
Unpubl. MA thesis. San Francisco State Univ., San Francisco, CA.
. 1982. Taxonomy and systematics of the velvetfishes (Pisces:
Aploactinidae). Unpubl. PhD dissertation. Univ. Michigan, Ann
Arbor.
. and W. N. Eschmeier. 1975. The Indo-Wesl Pacific scor-
pionfish genus Ocosia Jordan and Starks (Scorpaenidae, Tetraro-
ginae), with descriptions of three new species. Matsya 1:1-18.
, AND . 1978. Two new Australian velvetfishes, genus
Paraploactis (Scorpaeniformes: Aploactinidae), with a revision of
the genus and comments on the genera and species of the Aploac-
tinidae. Proc. Calif Acad. Sci. 41:401-426.
Post, A. 1976. Ergebnisse der Forschungreisen des FFS "Walther
Herwig" nach Siidamenka. XLII. Direlmus Johnson. 1863 (Beryci-
formes, Berycoidei, Diretmidae). 2. Morphologie, Entwicklung,
Verbreitung. Arch. Fischereiwiss. 26(2/31:87-114.
, andJ. QiiERO. 1981. Revision des Diretmidae (Pisces, Trach-
ichthyoidei) de L'Atlantique avec description d'un nouveau genre
et d'une nouvelle espece. Cybium, 3rd Ser. 5(l):33-60.
PosTUMA. K. H. 1974. The nucleus of the herring otolith as a racial
character. J. Cons. Int. Explor. Mer 35:121-129.
POTTHOFF, T. 1974. Osteological development and variation in young
tunas, genus Thunnus (Pisces, Scombridae) from the .Atlantic Ocean.
Fish. Bull. U.S. 72:563-588.
. 1975. Development and structure of the caudal complex, the
vertebral column, and the pterygiophores in the blackfin tuna I'hun-
nus atlanlicus (Pisces. Scombndae). Bull. Mar. Sci. 25:205-231.
. 1980. Development and structure of fins and fin supports in
dolphin fishes Corvphaena hippuru.^ and Corvphaena equisetis
(Coryphaenidae). Fish. Bull. U.S. 78:277-312.
, AND S. ICellev. 1982. Development of the vertebral column,
fins and fin supports, branchiostegal rays, and squamation in the
swordfish, Xiphias gladius. Fish. Bull. U.S. 80:161-186.
, andJ.Javech. MS. Comparisons of the development
of the vertebral column, fins and fin supports in scombroid fishes.
, , M. MOE and F. Young. 1984. Description of porkfish
larvae (.-inisotremus virginicus. Haemulidae) and their osteological
development. Bull. Mar. Sci. 34:21-59.
, W. J. Richards and S. Ueyanagi. 1980. Development of
Scombrolabrax helerolepis (Pisces, Scombrolabracidae) and com-
ments on familial relationships. Bull. Mar. Sci. 30:329-357.
PowLES, H. 1977. Description of larval Jenkinsia lamprolaema (Clu-
peidae, Dussumieriinae) and their distribution off Barbados, West
Indies. Bull. Mar. Sci. 27:788-801.
. 1980. Descriptions of larval silver perch, Ba;>(//e//ac/irv50Mra.
banded drum, Lanmus fascialus, and star drum, Steltifer lanceo-
latus (Sciaenidae). Fish. Bull. U. S. 78:1 19-136.
, and B. W. Stender. 1978. Taxonomic data on the early life
history stages of Sciaenidae of the South Atlantic Bight of the United
States. S. C. Mar. Resour. Cent. Tech. Rep. 31.
Premalatha, P. 1977. A study of the development and distribution
of the larvae of leather skin Chonnemus sanctipetn (Cuv. & Val.)
(Carangidae-Pisces) along the southwest coast of India. Proc. Symp.
warm water zooplankton Spec. Publ. UNESCO/NIO:450-459.
Presley, R. F. 1970. Larval snowy grouper, Epincphetus niveatus
(Valenciennes, 1828), from the Florida straits, Fla. Dep. Nat. Re-
sour. Mar. Res. Lab. Leafl. Ser. 4(18).
PROKis, M.. and M. Pen.az. 1980. Early development of the chub.
Leuciscus cephahis. Acta Sci. Nat. Acad. Sci. Bohemoslov Brno
14(7): 1-40.
Qasim, S. Z. 1956. The spawning habits and embryonic development
of the shanny (Blenmus pholis L.). Proc. Zool. Soc. Lond. 127:79-
93.
QuAST, J. C. 1965. Osteological characteristics and affinities of the
hexagrammid fishes, with a synopsis. Proc. Calif Acad. Sci. 31:
563-600.
QuERo, J. 1978. Repartition des Zeides (Pisces, Zeiformes) captures
dans I'Atlantique oriental entre le Cap Vert (15°N) et le Cap Juby
(28°N) au cours des campagnes de la "Thalassa." Bull. Off Natl.
Pech. Tunis. 1978, 2(l-2):49-6l.
. 1979. Remarques sur le Grammicolepis hrachiusculus (Pisces,
Zeiformes, Grammicolepididae) observe au port de la Rochelle.
Ann. Soc. Sci. Nat. Charente-Marit. 6:573-576.
QiTNN, R. A.. AND C. C. SiGL (eds.). 1980. Radiography in modem
industry. Eastman Kodak Co., Rochester, NY.
Rabalais, N. N., S. C. Rabalais andC. R. Arnold. 1980. Description
of eggs and larvae of laboratory-reared red snapper (Lutjanus cam-
pechanus). Copeia 1980:704-708.
Rachow, a. 1924. Uber Cypnnodon iherus Valenciennes und Fun-
dulus hispanicus (Valenciennes). Blatt Aquar.-Terrarienk. 22(21):
369-371.
Radtke, R. L. 1984. Otolith formation and increment deposition in
laboratory reared skipjack tuna, Kalsuwomis pelanus larvae. NOA.^
Tech. Rep. In Press.
, AND J. M. Dean. 1982. Increment formation in the otoliths
of embryos, larvae and juveniles of the mummichog, Fundulu.'i
heterodilus. Fish. Bull. U.S. 80:201-216.
, AND K. G. Waiwood. 1980. Otolith formation and body
shrinkage due to fixation in larval cod (Gadus morhua). Can. Tech.
Rep. Fish. Aquat. Sci. 929.
Raffaele, F. 1888. Le uova galleggianti e le larve dei Teleostei nel
golfo di Napoli. Mitt. Zool. Sta. Neapel 8.
R AjAGOPAL, P. K. 1 979. The embryonic development and the thermal
effects on the development of the mountain whitefish, Frnsopium
williamsoni (Girand). J. Fish Biol. 15:153-158.
Raju, S. N. 1974. Three new species of the genus A/ono.gnaf/iK.'; and
the leptocephali of the order Saccopharyngiformes. Fish Bull. U.S.
72:547-562.
730
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Randall, J. E. 1955a. An analysis of the genera of surgeon fishes
(family Acanthuridae). Pac. Sci. 9:359-367.
. 1955b. A revision of the surgeonfish genera Ze/iraiOHia and
Paracanthurus. Pac. Sci. 9:396-412.
. 1 955c. A revision of the surgeonfish genus Ctenochaetus. fam-
ily Acanthuridae, with descriptions of five new species. Zoologica
(N.Y.) 40:149-166.
. 1956. A revision of the surgeon fish genus Acanthurus. Pac.
Sci. 10:159-235.
. 1961. A contribution to the biology of the convict surgeonfish
of the Hawaiian Islands, Acanthurus triostegus sandvicensis. Pac.
Sci. 15:215-272.
. 1980. Revision of the fish genus Plectranthias (Serranidae:
Anthiinae) with descriptions of 13 new species. Micronesica 16:
101-187.
. 1981. Underwater guide to Hawaiian reef fishes. Harrowood
Books, Newton Square, PA.
, AND R. Lubbock. 1981. A revision of the serranid fishes of
the subgenus Mirolabhchthys (Anthiinae: Anthias), with descrip-
tions of five new species. Contrib. Sci. (Los Ang.) 333.
, AND J. E. McCosKER. 1975. The eels of Easter Island with a
description of a new moray. Contrib. Sci. (Los Ang.) 264.
— , AND H. Randall. 1963. The spawning and early development
of the Atlantic parrot fish. Sparisoma rubripinne, with notes on
other scarid and labrid fishes. Zoologica (N.Y.) 48(2):49-60.
— , T. Shimizu AND T. Yamakawa. 1982. A revision of the Hol-
ocentrid fish genus Ostichthys, with descriptions of four new species
and a related new genus. Jpn. J. Ichthyol. 29:1-26.
Ranzi, S. 1933. Famiglia 3: Sparidae, p. 332-382. In: Uova, larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
Rao, K. S. 1973. Eggs and early developmental stages of Hilsa kelee
(Cuvier). Indian J. Fish. 20:250-255.
Rao, R. T. 1977. Enhancement of natural populations of moi (Po/r-
dactylus sextilis) in Hawaii through the release of hatchery-reared
juveniles— a feasibility study of sea ranching. Hawaii Inst. Mar.
Biol. Tech. Rep. 33.
Rao, V. V. 1970. Breeding habits and early development of two blen-
niid fishes Omobranchus japontcus (Bleeker) and Cruantus smithi
(Visweswara Rao) from Godavari Estuary. J. Mar. Biol. Assoc.
India 12(1&2):175-182.
Rapson, a. M. 1940. The reproduction, growth, and distnbution of
the lemon soles [Pelotretis ftavilatus Waite) of Tasman Bay and
Marlborough Sounds. N.Z. Mar. Dep. Fish. Bull. 7:1-56.
Rasquin, p. 1958. Ovarian morphology and early embryology of the
pediculate fishes Anlennarius and Histno. Bull. Am. Mus. Nat.
Hist. 1 14:327-372.
Rass, T. S. 1949. Composition of the ichthyoplankton of the Barents
Sea. In: T. S. Rass, I. I. Kazanova, S. P. Alekseyeva and L. A.
Ponomareva (eds.). Material on the reproduction and development
of fishes of the Northern Seas. Tr. Vses. Nauchno-Issled. Inst. Morsk.
Rybn. Khoz. Okeanogr. 17:1-68.
. 1950. An unusual instance of a biological relation between
fish and crabs. Prioroda 39(7):68-69.
. 1954. Contribution to the study of Pacific Ocean Moridae
(Pisces, Gadiformes). Tr. Inst. Okeanol. Akad. Nauk. SSSR 1 1:56-
61.
. 1 960. On the geographical distribution of bathypelagic fishes
of the family Myctophidae in the Pacific Ocean. Tr. Inst. Okean.
Akad. Nauk SSSR 41:146-152. (In Russian.)
. 1968. Spawning and development of polar cod. Rapp. P. V.
Reun. Cons. Perm. Int. Explor. Mer 158:135-137.
. 1972. Ichthyoplankton and the deep-sea fishes of the world
ocean. Trans. P. P. Shirshov. Inst. Oceanol. Acad. Sci. USSR 93:
34-35. (In Russian.)
. 1973. Systematics of pelagic fish eggs and larvae (Ichthyo-
plankton). Unpubl. manuscript. FAO, Rome.
Rathbun, R. 1893. Report upon the inquiry respecting food fishes
and the fishing grounds. Rep. U.S. Comm. Fish. 17:97-171.
Ray, D. 1950. The peripheral nervous system of Lampanyctus leu-
copsarus. J. Morphol. 87:61-178.
Re, P. 1980. The eggs and newly hatched larvae of /liu(^e/(/M//!^nrf«i
(Cuvier, 1 830) (Pisces: Pomacentridae) from the Azores. Arq. Mus.
Bocage, 2nd Ser. 7(8):109-1 16.
Reay, P. J. 1970. Synopsis of biological data on North Atlantic sand-
eels of the genus Ammodytes. A. lobianus, A. dubius. A. aniericanus.
and A. mannus. FAO Fish. Synop. 82, Rome.
Regan, C. T. 1902. On the systematic position of Luvarus imperialis
Rafinesque. Ann. Mag. Nat. Hist., Ser. 7, 10:278-281.
. 1903a. Descriptions de poissons nouveaux faisant partie de
la collection du Musee d'histoire naturelle de Geneve. Rev. Suisse
Zool. 11:413-418.
. 1903b. LXII. — On the systematic position and classification
of the gadoid or anacanthine fishes. Ann. Mag. Nat. Hist., Ser. 7,
11:459-466.
-. 1907. On the anatomy, classification and systematic position
of the teleostean fishes of the suborder Allotriognathi. Proc. Zool.
Soc. Lond. 1907:634-643.
. 1909. On the anatomy and classification of the scombroid
fishes. Ann. Mag. Nat. Hist., Ser. 8, 3:66-75.
. 1910. The origin and evolution of the teleostean fishes of the
order Heterosomata. Ann. Mag. Nat. Hist., Ser. 8, 6:484-496.
. 1911a. The anatomy and classification of the teleostean fishes
of the order Iniomi. Ann. Mag. Nat. Hist., Ser. 8, 7:120-133.
. 1911b. The classification of the teleostean fishes of the order
Synentognathi. Ann. Mag. Nat. Hist., Ser. 8, 7:327-335.
. 1911c. The osteology and classification of the gobioid fishes.
Ann. Mag. Nat. Hist., Ser. 8, 8:729-733.
. 1912a. The classification of the teleostean fishes of the order
Pediculati. Ann. Mag. Nat. Hist.. Ser. 8, 9:277-289.
. 1912b. The classification of the blennioid fishes. Ann. Mag.
Nat. Hist., Ser. 8, 10:265-280.
. 1912c. The osteology and classification of the teleostean fishes
of the order Apodes. Ann. Mag. Nat. Hist., Ser. 8, 10:377-387.
. 1913a. Osteology and classification of the teleostean fishes of
the order Scleroparei. Ann. Mag. Nat. Hist., Ser. 8, 1 1:169-184.
. 1913b. On the classification of the percoid fishes. Ann. Mag.
Nat. Hist., Ser. 8, 12:111-145.
. 1 9 1 3c. Phallostethus dunckert. a remarkable new cyprinodont
fish from Johore. Ann. Mag. Nat. Hist., Ser. 8, 12:548-555.
. 1913d. The Antarctic fishes of the Scottish National Antarctic
Expedition. Trans. R. Soc. Edinb. 49(2):229-292.
. 1916. Larval and post-larval fishes. Nat. Hist. Rep. Br. Ant-
Terra Nova Exped. Zool. 1:125-156.
1923. The classification of the stomiatoid fishes. Ann, Mag.
Hist., Ser. 9, 11:612-614.
. 1924. The morphology of a rare oceanic fish, Stylophorus
chordatus Shaw; based on specimens collected in the Atlantic by
the "Dana" expeditions, 1920-1922. Proc. R. Soc. Lond. 96:193-
207.
. 1925. III. The fishes of the genus G/gflWwra, A. Brauer; based
on the specimens collected in the Atlantic by the "Dana" Expe-
ditions, 1920-22. Ann. Mag. Nat. Hist. 15(9):53-59.
. 1926. The pediculate fishes of the suborder Ceratioidea.
Oceanogr. Rep. Dana Exped. 2.
. 1929. Fishes. Encyclopaedia Britannica, 14th ed. 9:305-328.
, AND E. Trewavas. 1929. The fishes of the families Astrones-
thidae and Chauliodontidae. Dana-Rep. Carlsberg Found. 5.
, AND . 1930. The fishes of the families Stomiidae and
Malacosteidae. Dana-Rep. Carlsberg Found. 6.
.AND . 1932. Deep-sea angler-fishes (Ceratioidea). Dana-
Rep. Carlsberg Found. 2.
Renzhai, Z., AND L. SuiFEN. 1980a. Studies on the eggs and larvae
development of wise porgy, Gymnocranius griseus (Temminck et
Schlegel). Acta Zool. Sin. 26(2): 132-1 35. (In Chinese, English ab-
stract.)
, and . 1980b. On the eggs and larvae of the golden-
arct.
Nat.
LITERATURE CITED
731
thread Nemiptenis virgalus (Houttuyn). Acta Zool. 26(2): 136- 139.
In Chinese. English abstract.)
Reshetnikov, Yu. S. 1975. The systematic position of whitefishes.
Zool. Zh. 54:1651-1671. (In Russian.)
Richards, S. W. 1965. Description of the posllarvae of the sand lance
(Ammodyles) from the east coast of North America. J. Fish. Res.
Board Can. 22:1313-1317.
. 1982. Aspects of the biology of Ammodyles americanus from
the St. Lawrence River to Chesapeake Bay, 1972-75, including a
comparison of the Long Island Sound postlarvae with Ammodytes
dubius. J. Northwest Atl. Fish. Sci. 3:93-104.
Richards, W. J. 1969. Elopoid leptocephali from Angolan waters.
Copeia 1969:515-518.
. 1974. Evaluation of identification methods for young billfish-
es, p. 62-72. In: Proc. Intematl. Billfish Symp., Kailua-Kona, Ha-
waii, 9-12 August 1972, Part 2. Review and contributed papers.
R. S. Shomura and F. Williams (eds.).
. 1981. Kinds and abundances of fish larvae in the Caribbean
Sea. Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:240-241.
. 1984. Kinds and abundances of fish larvae in the Caribbean
Sea and adjacent areas. NOAA Tech. Rep. NMFS SSRF-776.
, AND G. R. Dove. 1971. Internal development of young tunas
of the genera Katsuwonus. Euthynnus. Auxis. and Thunnus (P'xscti,
Scombndae). Copeia 1971:72-78.
, AND W. L. Klawe. 1972. Indexed bibliography of the eggs
and young of tunas and other scombrids (Pisces, Scombridae) 1 880-
1970. NOAA Tech. Rep. NMFS SSRF-652.
, R. V. Miller and E. D. Houde. 1974. Eggs and larval de-
velopment of the Atlantic thread herring, Opisthonema ogtimun.
Fish. Bull. U.S. 72:1123-1136.
, AND T. PoTTHOFF. 1974. Analysis of the taxonomic characters
of young scombrid fishes, genus Thunnus. p. 623-648. In: The
early life history of fish. J. H. S. Blaxter (ed.). Springer- Verlag,
Berlin.
-, AND V. P. Saksena. 1980. Description of larvae and early
juveniles of laboratory-reared gray snapper, Luljanus griseus (Lin-
naeus) (Pisces, Lutjanidae). Bull. Mar. Sci. 30:515-521.
Richardson, S. L. 1981a. Current knowledge of larvae of sculpins
(Pisces: Cottidae and allies) in northeast Pacific genera with notes
on intergeneric relationships. Fish. Bull. U.S. 79:103-121.
. 1981b. Pelagic eggs and larvae of the deepsea sole, Embas-
sichthys bathybius (Pisces: Pleuronectidae), with comments on ge-
neric affinities. Fish. Bull. U.S. 79:163-170.
. 1981c. Current knowledge of northeast Pacific sculpin larvae
(Pisces: Cottidae and allies) with notes on relationships within the
group. Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:603-604.
, AND D. A. Dehart. 1975. Records of larval, transforming,
and adult specimens of the quillfish, PtiUchlhys goodei from waters
off Oregon. Fish. Bull. U.S. 73:681-684.
, J. R. Dunn andN. A. Naplin. 1980. Eggs and larvae of butter
sole, Isopsetta isolepis (Pleuronectidae), off Oregon and Washing-
ton. Fish. Bull. U.S. 78:401-417.
, and E. B. Joseph. 1973. Larvae and young of western North
Atlantic bothid flatfishes Etropus microslomiis and Cilhanchlhys
arctifrons in the Chesapeake Bight. Fish. Bull. U.S. 71:735-767.
, and W. A. Laroche. 1979. Development and occurrence of
larvae and juveniles of the rockfishes, Sebastes crameri. Sebastes
pinniger and Sebasles helvomaculalus (Family Scorpaenidae) off
Oregon. Fish. Bull. U.S. 77:1-41.
, and W. G. Pearcy. 1977. Coastal and oceanic fish larvae in
an area of upwelling off Yaquina Bay, Oregon. Fish. Bull. U.S. 75:
125-145.
, AND B. B. Washington. 1 980. Guide to identification of some
sculpin (Cottidae) larvae from marine and brackish waters off Or-
egon and adjacent areas in the northeast Pacific. NOAA Tech. Rep.
NMFS Circ. 430.
Riehl, R. 1980. Micropyle of some salmonins and coregonins. En-
viron. Biol. Fishes 5:59-66.
RiLE'i, J. D. 1966. Marine fish culture in Britain, VII. Plaice {Pteu-
ronectes plalessa L.) post-larval feeding on Anemia salina L. napulii
and the effects of varying feeding levels. J. Cons. Int. Explor. Mer
30:204-221.
RiTTER, W. E., AND S. E. Bailey. 1908. On the weight of developing
eggs. Univ. Calif Publ. Zool. 6:1-10.
RivAS, L. R. 1972. Opisthonema captivai. a new western Atlantic
clupeid fish from Columbia. Copeia 1972:1-3.
, AND S. M. Warlen. 1967. Systematics and biology of the
bonefish, Albula nemoptera (Fowler). U.S. Fish Wildl. Serv. Fish.
Bull. 66(2):25 1-258.
Roberts, R. J., M. Bell and H. Young. 1973. Studies on the skin
of plaice (Plenronectes platessa L.). II. The development of larval
plaice skin. J. Fish Biol. 5:103-108.
Roberts, T. R. 1970. Description, osteology and relationships of the
Amazonion cyprinodont fish Fhiviphylax pygmaeus (Myers and
Carvalho). Breviora 347:1-28.
. 1971. The fishes of the Malaysian family Phallostethidae(Ath-
eriniformes). Breviora 374:1-27.
. 1981. Sundasalangidae, a new family of minute freshwater
salmoniform fishes from southeast Asia. Proc. Calif. Acad. Sci. 42:
295-302.
. 1982a. The southeast Asian freshwater pufferfish genus Cho-
nerhinos (Tetraodontidae), with descnptions of new species. Proc.
Calif Acad. Sci. 43:1-16.
. 1982b. Unculi (homy projections arising from single cells),
an adaptive feature of the epidermis of ostariophysan fishes. Zool.
Scr. ll(l):55-76.
Robertson, D. A. 1973. Planktonic eggs and larvae of some New
Zealand marine teleosts. Unpubl. PhD thesis. Univ. Otago, Dune-
din, N.Z.
. 1974. Developmental energetics of the southern pigfish (Tele-
ostei: Congiopodidae). N.Z. J. Mar. Freshwater Res. 8:61 1-620.
. 1975a. A key to the planktonic eggs of some New Zealand
marine teleosts. N.Z. Fish. Res. Div. Occas. Publ. No. 9.
. 1975b. Planktonic stages of the teleost family Carapidae in
eastern New Zealand waters. N.Z. J. Mar. Freshwater Res. 9:403-
409.
. 1976. Planktonic stages of Mauroticus muelleri (Teleostei:
Stemoptychidae) in New Zealand waters. N.Z. J. Mar. Freshwater
Res. 10:311-328.
. 1977. Planktonic eggs of the lantemfish Lanzpawctoo'e.s /lec-
Mw (family Myctophidae). Deep-Sea Res. 24:1-4.
. 1978. Spawning of the tarakihi (Pisces: Cheilodactylidae) in
New Zealand waters. N.Z. J. Mar. Freshwater Res. 12:277-286.
. 1981. Possible functions of surface structure and size in some
planktonic eggs of manne fishes. N.Z. J. Mar. Freshwater Res. 15:
147-153.
, AND S. Mito. 1979. Sea surface ichthyoplankton off south-
eastern New Zealand, summer 1977-78. N.Z. J. Mar. Freshwater
Res. 13:415-424.
, AND U. Raj. 1971. Egg and yolk sac stages of the sand flounder
(Teleostei: Pleuronectidae). N.Z. J. Mar. Freshwater Res. 5:404-
414.
, P. E. Roberts and J. B. Wilson. 1978. Mesopelagic faunal
transition across the Subtropical Convergence east of New Zealand.
N.Z. J. Mar. Freshwater Res. 12:295-312.
Robertson, D. R. 1973. Field observations on the reproductive be-
haviour of a pomacentrid fish, Acanlhochrnmis polyacanthus. Z.
Tierpsychol. 32:319-324.
Robins, C. H., and C. R. Robins. 1976. New genera and species of
dysommine and synaphobranchine eels (Synaphobranchidae) with
an analysis of the Dysomminae. Proc. Acad. Nat. Sci. Phila. 127:
249-280.
Robins, C. R. 1958. Studies on fishes of the family Ophidiidac. I. A
new species of Lepophidium from the Caribbean Sea. Bull. Mar.
Sci. Gulf Caribb. 8:360-368.
732
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
. 1959. Studies on fishes of the family Ophidiidae. III. A new
species of Lepophidium from Barbados. Breviora 104.
— . 1960. Studies on fishes of the family Ophidiidae. V. Lepo-
phidium pherornystax, a new Atlantic species allied to Lepophidium
jeannae Fowler. Bull. Mar. Sci. GulfCaribb. 10:83-95.
. 1961. Studies on fishes of the family Ophidiidae. VI. Two
new genera and a new species from American waters. Copeia 1961:
212-221.
. 1962. Studies on fishes of the family Ophidiidae. VII. The
Pacific species of Lepophidium. Copeia 1962:487-498.
. 1967. The status of the serranid fish Liopropoma aberrans
with the description of a new, apparently related genus. Copeia
1967:591-595.
. 1974. The validity and status of the roundscale spearfish.
Telrapturus georgei. p. 54-61. In: Proc. Int. BiUfish Symposium
Kailua-Kona, Hawaii, 9-12 August 1972. Part 2. Review and Con-
tributed Papers. R. S. Shomuraand F. Williams (eds.). NOAATech.
Rep. NMFS SSRF-675.
, R. M. Bailey, C. E. Bond, J. R. Brooker, E. A. Lachner, R.
N. Lea AND W. B. Scott. 1980. A list of common and scientific
names of fishes from the United States and Canada (Fourth Edi-
tion). Am. Fish. Soc. Spec. Publ. 12.
, AND J. E. BoHLKE. 1959. Studies On fishcs of the family Ophi-
diidae. IV. Two new dwarf cusk-eels (genus Ophidian) from the
tropical western Atlantic. Not. Nat. (Phila.) 325.
— , AND . 1970. The first Atlantic species of the ammodytid
fish genus Embolichlhys. Not. Nat. (Phila.) 430:1-1 1.
, AND D. P. DE Sylva. 1960. Description and relationships of
the longbill spearfish, Tetrapierus betone. based on western North
Atlantic specimens. Bull. Mar. Sci. Gulf and Caribb. 10:383-413.
— , AND . 1965. The Kasidondae. a new family of mira-
pinniform fishes from the western Atlantic Ocean. Bull. Mar. Sci.
15:189-201.
— , AND R. N. Lea. 1978. Fishes of the family Ophidiidae. IX.
Analysis of geographic variation in Lepophidium microlepis with
descriptions of two new subspecies. Bull. Mar. Sci. 28:716-727.
— , AND J. G. Nielsen. 1 970. Snydendia bothrops, a new tropical,
amphi-Atlantic species (Pisces: Carapidae). Stud. Trop. Oceanogr.
(Miami) 4(2):285-293.
— , AND J. E. Randall. 1965. Three new western Atlantic fishes
of the blennioid genus Chaenopsis. with notes on the related Lu-
cayablennius zingaro. Proc. Acad. Nat. Sci. Phila. 117:213-277.
RoFEN,R.R. 1958. ThemarinefishesofRennell Island, p. 149-218.
In: The natural history of Rennell Island British Solomon Islands.
Vol. I. Danish Science Press, Ltd., Copenhagen.
. 1966a. Family Paralepididae, p. 205-461. In: Fishes of the
western North Atlantic. G. W. Mead (ed.). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 5).
. 1966b. Family Omosudidae, p. 462-481. In: Fishes of the
western North Atlantic. G. W. Mead (ed.). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 5).
. 1966c. Family Anotopteridae, p. 498-410. /«. Fishes of the
western North Atlantic. G. W. Mead (ed.). Mem. Sears Found.
Mar. Res. No. 1 (Pt. 5).
. 1966d. Family Evermannellidae, p. 511-565. In: Fishes of
the western North Atlantic. G. W. Mead (ed.). Mem. Sears Found.
Mar. Res. No. I (Pt. 5).
Rohde, F. C. 1 980. Dallia pectoralis. p. 1 27. In: Atlas of North Amer-
ican freshwater fishes. D. S. Lee, C. R. Gilbert, C. H. Hocutt, R.
E. Jenkins, D. E. McAllister and J. R. Stauffer, Jr. (eds.). North
Carolina State Mus. Nat. Hist.. Raleigh, NC.
Rose, J. A. 1961. Anatomy and sexual dimorphism of the swim blad-
der and vertebral column in Ophidian holbrooki (Pisces: Ophidi-
idae). Bull. Mar. Sci. Gulf Canbb. 1 1:280-308.
Rosen, D. E. 1964. The relationships and taxonomic position of the
halfbeaks, killifishes, silversides, and their relatives. Bull. Am. Mus.
Nat. Hist. 127:217-268.
. 1971. The Macristiidae, a ctenothrissiform family based on
juvenile and larval scopelomorph fishes. Am. Mus, Nov. 2452.
— . 1973. Intertelationships of higher euteleostean fishes, p. 397-
513. In: Interrelationships of fishes. P. H. Greenwood. R. S. Miles,
and C, Patterson (eds.). Zool. J. Linn. Soc. 53 (Suppl. 1.).
. 1974. Phylogeny and zoogeography of salmoniform fishes and
relationships of Lepidogala.xias satamandroides. Bull. Am. Mus.
Nat. Hist. 153:267-325.
. 1978. Vicariant patterns and historical explanation in bio-
geography. Syst. Zool. 27:159-188.
. 1982. Teleostean interrelationships, morphological function
and evolutionary inference. Am. Zool. 22:261-273.
. 1983. Zeoidsare tetraodontiforms, acanthuridsare not. ASIH
Program & Abstracts, 1983.
. 1984. Zeiforms as primitive plectognath fishes. Am. Mus.
Novit. 2782.
— , and p. H. Greenwood. 1970. Origin of the Weberian ap-
paratus and the relationships of the ostariophysan and gonoryn-
chiform fishes. Am. Mus. Novit. 2428.
— , AND L. R. Parenti. 1981. Relationships of On'r/fli. and the
groups of atherinomorph fishes. Am. Mus. Novit. 2719.
— , AND C. Patterson. 1969. The structure and relationships of
the paracanthopterygian fishes. Bull. Am. Mus. Nat. Hist. 141:357-
474.
Rosenblatt, R. H. 1959. A revisionary study of the blennioid fish
family Tripterygiidae. Unpubl. PhD dissertation. Univ. Calif, Los
Angeles.
. 1960. The Atlantic species of the blennioid fish genus En-
neanectes. Proc. Acad. Nat. Sci. Phila. 1 12(1): 1-23.
, AND M. A. Bell. 1976. Osteology and relationships of the
roosterfish, Nematistius pectoralis Gill. Contrib. Sci. (Los Ang.)
279:1-23.
, and J. L. Butler. 1977. The ribbonfish genus Z)Mmorfe»ia,
with the description of a new species (Pisces, Trachipteridae). Fish.
Bull. U.S. 75:843-855.
, AND G. D. Johnson. 1976. Anatomical considerations of pec-
toral swimming in the opah, Lampns gutlatus. Copeia 1976:367-
370.
, AND T. D. Parr. 1969. The Pacific species of the clinid fish
genus Paraclinus. Copeia 1969:1-20.
, AND J. S. Stephens. 1978. Mccoskerichthys sandae. a new and
unusual chaenopsid blenny from the Pacific coast of Panama and
Costa Rica. Contnb. Sci. (Los Ang.) 293:1-21.
, AND L. R. Tailor, Jr. 1971. The Pacific species of the clinid
fish tribe Starksiini. Pac. Sci. 25:436-463.
Rosenthal, H. 1969. Verdauungsgeschwindigkeit, Nahrungswahl und
Nahrungsbedarf bei den Larven des Herings, Clupea harengus L,
Ber. Dtsch. Wiss. Komm. Meeresforsch. 20:60-69.
, and G. Hempel. 1970. Experimental studies in feeding and
food requirements of herring larvae (Clupea harengus L.), p. 344-
364. In: Marine food chains. J. H. Steele (ed.). Univ. Calif Press,
Berkeley.
Rostand, J. 1964. Biological and geological sciences. Chapt. 1. Gen-
eral biology. In: The beginnings of modem science. Basic Books,
New York.
Roule, L. 1922. Dtscnplionde Scombrolabrax heterolepis noy . gen.
nov. sp., poisson abyssal nouveau de I'lle Madere. Bull. Inst. Ocean-
ogr. (Monaco) 408.
. 1924. Etude sur Pontogenese et al croissance avec hypermet-
amorphose de Luvarus imperialis Raf (poisson rapporte a Tordre
des Scombriformes). Ann. Inst. Oceanogr. Monaco (n.s.) 1:119-
157.
, AND F. Angel. 1 930. Larves et alevins de poissons provenant
des croisieres du Prince Albert I de Monaco. Result. Camp. Sci.
Prince Albert I, 79:1-148.
RoxAS, H. A. 1934. A review of Philippine Mugilidae. Philipp. J. Sci.
54(3):393-429.
Ruck, J. G. 1971. Development of the lumpfish, Trachelochismus
melobesia (Pisces: Gobiesocidae). Zool. Publ. Victoria Univ. Wel-
lington 57:1-9.
. 1973a. HeyeXopmexW. of Tripier\-gion capito 2iTid T. robustum
LITERATURE CITED
733
(Pisces: Tripterygiidae). Zool. Publ. Victoria Univ. Wellington 63:
1-10.
. 1973b. Developmeul of ihe clingfishes, Diplocrepis puniceus
and Trachelochismiis pmnii/alus (Pisces: Gobiesocidae). 2tool. Publ.
Victoria Univ. Wellington 64:1-12.
. 1976. Studies on the development and osteology of some New
Zealand inshore fishes. Unpubl. PhD dissertation. Victoria Univ.
Wellington, New Zealand.
. 1980. Early developmenl of Forslery-gion vanum. Gilloblen-
nius decemdigitatus. and G. tnpennis (Pisces: Tripterygiidae). N. Z.
J. Mar. Freshwater Res. 14:313-326.
RuDDELL, C. L. In press. Initiating polymerization of glycol methac-
rylate with cyclic diketo carbon acids. Stain Technol.
RuDOMETKiNA, G. P. 1980. Larvae oflizardfish (family Synodontidae)
from the eastern tropical Atlantic. J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 20:(3):147-152.
. 1981. Postembryonic development of Diplophos taenia Gun-
ther, 1873 (Gonostomatidae) from the .Atlantic Ocean. J. Ichthyol.
(Engl. Transl. Vopr. Ikhtiol.) 21(4):I06-1 13.
RuNYAN, S. 1961. Early development of clingfish, Gobiesox strumosus
Cope. Chesapeake Sci. 2:113-141.
RupLE, D. MS. Larvae of some gobioid fishes from the Western North
Atlantic.
Russell, B. C. 1980. Revision of the fish genus Pseudolabrus and
allied genera (family Labridae) with a phylogenetic analysis of re-
lationships. Unpubl. PhD dissertation. Macquarie Univ., Sydney.
Russell, F. S. 1976. The eggs and planktonic stages of British marine
fishes. Academic Press, London.
Russo,J. L. 1983. Interrelationshipsof the gempylid fishes (Teleostei.
Scombroidei). Unpubl. PhD dissertation. George Washington Univ.,
Washington, D.C.
RuTENBERG, E. P. 1962. Obzor ryb semeistva terpugovykh (Hexa-
grammidae) (Survey of the fishes of family Hexagrammidae). Tr.
Inst. Okeanol. Akad. Nauk SSSR 59:3-100. (Transl. Isr. Prog. Sci.
Transl., 1970, p. 1-103. In: Greenlings: Taxonomy, biology, in-
teroceanic transplantation. T. S. Rass (ed). Avail.: NTIS. Spring-
field, VA; TT69-55097.)
Ryder, J. A. 1883. Observations on the absorption of the yolk, the
food, feeding, and development of embryo fishes, comprising some
investigations conducted at the central hatchery. Armory Building,
Washington D.C. in 1882. Bull. U.S. Fish. Comm. 2( 1882): 179-
205.
. 1884. A contribution to the embryography of osseous fishes,
with special reference to the cod(Gadus morhua). Rep. U.S. Comm.
Fish. 10(1882):455-605.
. 1885. On the development of viviparous osseous fishes and
of the Atlantic salmon. Proc. U.S. Natl. Mus. 8:128-162.
. 1887. On the development of osseoiis fishes, including marine
and freshwater forms. Rep. U.S. Comm. Fish. 13(1885):489-603.
. 1888. Development of the sea-bass (Serra/iiw a?rani«). Am.
Nat. 22:755.
Saf'Yanova, T. Y., and N. I. Revina. 1960. The biology of the large
horse mackerel and fishing for it. Tr. Azovo-Chemomorsk. Nauch-
no-Issled. Inst. Mosrsk. Rybn. Khoz. Okeanogr. 18.
Saioal, B. N., AND M. P. MoTWANL 1962. Studies on the fishery and
biology of the commercial cat-fishes of the Ganga River system. 1.
Early life-history, bionomics and breeding of Mystus (Osteobagrus)
seenghala Sykes. Indian J. Fish. 8(I):60-74.
Sakamoto, T., AND K. Suzuki. 1978. Spawning behavior and early
life history of the porcupine puffer, Diodon holacanthus. in aquaria.
Jpn. J. Ichthyol. 24:261-270.
Saksena, V. P., AND W. J. Richards. 1975. Description of eggs and
larvae of laboratory-reared white grunt, Haemulon plumien (La-
cepede) (Pisces: Pomadasyidae). Bull. Mar. Sci. 25:523-526.
Saldanha, L., and N, C. Pereira. 1977. Lophoius lacepedei Gioma,
1809 (Pisces, Lophotidae) signale pour la premiere fois sur la cote
portugaise. Arq. Mus. Bocage 6(4):51-59.
Salekhova, L. p. 1959. O razvitii lufarya (Pomawmus saltalrix Linne).
Tr. Sevastop. Biol. Stn. Im. A. D. Kovalenskogo Acad. Nauk Ukr.
SSR 11:182-188.
Sanchez Delgacxj, F. 1981. Contribucion al conocimiento de los
Labridos (Familia Labridae) de la costas Ibericas. Parte 1. Descrip-
cion de las especies. Bol. Inst. Esp. Oceanogr. 6(3): 19-57.
Sandercock, F. K., and N. J. Wilimovsky. 1968. Revision of the
cottid genus Enophrys. Copeia 1968:832-853.
Santander, H., and O, S. de Castillo. 1969. Desarrollo y distri-
bucion de huevos y larvas de merluza, Merluccius gayi (Guichenot)
en la costa peruana. Bol. Inst. Mar Peru (Callao) 2(3):79-107.
, and . 1971. Desarrollo y distribucion de huevos y larvas
de jurel ( Trachums symmetricus murphyi Nichols) en la costa peru-
ana. Inst. Mar Peru (Callao) Inf. 36:1-22.
Santerre, M. T., AND R. C. May. 1977. Some effects of temperature
on laboratory-reared eggs and larvae of Polydactylus sextilis (Pisces:
Polynemidae). Aquaculture 10:341-351.
Sanzo, L. 1911. Distnbuzione delle papille cutanee (Organi ciatiformi)
e suo valore sistematice nei Gobi. Mitt. Zool. Sta. Neapel 20:249-
328.
. 1912a. Larva di S'rom/a.sioa Risso. Mem. R. Com. Talassogr.
Ital. 10.
. 1912b. Uova di "Stemoptychidae" Gthr. R. Com. Talassogr.
Ital. Boll. Bimest. 19:384-386.
. 19 1 5a. Stadi larvali di Bathophilus nigerriinus Gigl. Mem. R.
Com. Talassogr. Ital. 48.
. 1915b. Stylopthalmoides lobiancoi Mazzarelli e S. mediter-
raneus Mazzarelli sono le rispettive forme larvali di Scopelus can-
imamis C. e V. e 5. humbohi Rissoi. R. Com. Talassogr. Ital. Boll.
44:1-24.
. 1917. Stadi larvali di /'ara/e;7« M'a//>Ja C.V. Mem. R. Com.
Talassogr. Ital. 59.
. 1918. Uova e larve di Trachipterus cristatits Bp. Mem. R.
Com. Talassogr. Ital. 64.
. 1925. \Jo\a.e\a.r\e (\\ Regalecus glesne Asc. Mem. K. Com.
Talassogr. Ital. 1 18.
. 1928. Uova e larve di Remora remora L. Mem. R. Com.
Talassogr. Ital. 138.
— . 1930a. Contributo alia conoscenza di uova, larve e stadi gio-
vanili in Echeneis naucrales Linn. Ann. Idrogr. Genova 11:201-
209.
— . 1930b. Uova, sviluppo embrionale. stadi larvali, post larvali
e giovanili di Stemoptychidae e Stomiatidae. Stemoptychidae. 2.
Ichthyococcus ovalus Cocco. R. Com. Talassogr. Ital. Monogr.
2:69-119.
— . 1930c. Contributo alia consoscenza della sviluppo nei Car-
ancidi: Seriola dumehlii Risso. Boll. Zool. 1:33-34.
— . 1930d. Ricerche biologiche su materiale raccolti dal Prof. L.
Sanzo nella campagna idiografica nei Mar Rosso della R.N. Am-
miraglio Magnaghi 1923-1924. VII. Plectognati. Mem. R. Com.
Talassogr. Ital. Monogr. 167.
— . 193 la. Uova stadi embrionali e post-embnonali di Naucrates
duclorL. Mem. R. Com. Talassogr. Ital. 185.
— . 1931b. Uova e larve di Zeus faber L. Arch. Zool. Ital. 15:
475.
. 1931c. Uova e primi stadi larvali di Myctophum gemellari
Cocco (=Scopelus gemellari C. e V.). Atti Accad. Naz. Lincei 14:
515-519.
— . 193 Id. Sottordine: Salmonoidei, p. 21-92. In: Uova, larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1932a. \Jo\ae X)nm\ ^ladWarvaM <X\ Trachurus meditenaneus.
Boll. Zool. 3:73.
— . 1932b. Uova, stadi larvali e giovanili di Centrolophus pom-
pilus C.V. Mem. R. Com. Talassogr. Ital. 196.
. 1933a. Famiglia 2: Macruridae, p. 255-265. //;.• Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38:255-265.
— . 1933b. Uova larve e stadi giovanili di 5er(o/a rfMmer//// Risso.
Mem. R. Com. Talassogr. Ital. 205.
734
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
— . 1 933c. Uova stadi larvali e giovanili di Dactylopterus voUtans
L. Mem. R. Com. Talassogr. Ital. 207.
— . 1936. Contributi alia conoscenza dello sviluppo embrionovio
nei Mugilidi. I. Uova e larve di Mugil cephalus Cuv. Ottenute per
tecondazione anificiale. II. Uova e larve di Mugil chelo Cuv. Mem.
R. Com. Talassogr. Ital. 230.
— . 1 938a. Uova, stadi embrionali, prelarve e larve di Saurenche-
lys canchvora Peters. Mem. R. Com. Talassogr. Ital. 249.
— . 1938b. Uova, ovariche e sviluppo larvale di Aulopus fila-
mentosus Cuv. Mem. R. Com. Talassogr. Ital. 254:3-7.
— . 1939a. Nuovo contributo alia conoscenza dello sviluppo di
Myctophum rissoi (Cocco). Atti Accad. Gioenia Sci. Nat. Catanias
(6)3(25): 1-8.
— . 1939b. Rarissimi stadi larvali di Teleostei. Arch. Zool. Ital.
26:121-151.
— . 1939c. Uova e larve di Smara (>!i/rf/arorC. V. Mem. R. Com.
Talassogr. Ital. 262.
. 1940. Uova e larve appena achiusa di Lophotes cepediamts
(Gioma). Mem. R. Com. Talassogr. Ital. 272.
1956. Divisione: Zeomorphi Regan, p. 461-470. In: Uova,
larve e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli
Monogr. 38.
Sardou, J. 1966. Oeufet developpement embryonnairede Trac/np-
terus taenia Bloch (ordre des Lampridiformes = AUotriognathes
famille des Trachypteridae) Vie Milieu, Fasc. 1-A, 15:199-215.
. 1974. Contribution a la connaissance de la fauna ichthyolo-
gique Malgache: decouverte de poissons de la famille des Schind-
leriidae dans le canal de Mozambique, a Nosy-Be et etude d'une
collection de Schindleria. Cah. ORSTOM, Ser. Oceanogr. 12(1):3-
15.
Sarojini, K. E., AND J. C. Malhotra. 1952. The larval development
of the so-called Indian salmon, Eleutheronema letradaclylum (Shaw).
J. Zool. Soc. India 4(l):63-72.
Sasaki, T. 1974. On the larvae of three species of rockfish (Genus:
Sebastes) in Hokkaido. Bull. Fac. Fish Hokkaido Univ. 25:169-
173.
Sato, G. 1983. Identifica?ao, distribuicao e desenvolvimento larval
de"Lanceta'" Thyrsitops lepidopoides (Cuvier, 1931) (Pisces: Gem-
pylidae) da regiao comprendida entre Cabo Frio (23°S) e Cabo de
Sta. Marta Grande (29°S). Unpubl. MS thesis. Instituto Oceano-
grafico da Universidade de Sao Paulo, Sao Paulo, Brazil. (In Por-
tuguese.)
Sato, M., and F. Yasuda. 1980. Metamorphosis of the leptocephali
of the ten-pounder, Elops hawaiensis. from Ishigaki Island, Japan.
Jpn. J. Ichthyol. 26:315-324.
Sauskan, V. I., AND V. P. Serebryakov. 1968. Reproduction and
development of the silver hake {Merluccius biltnearis Mitchill). J.
Ichthyol. (Engl. Trans. Vopr. Ikhtiol.) 8:398-414.
Saville, A. 1964. Clupeoidae. Identification sheets on eggs and larvae.
Sheet 1 . Cons. Int. Explor. Mer.
. 1981. The estimation of spawning stock size from fish egg and
larval surveys. Rapp. P. V. Reun. Cons. Int. Explor. Mer 178:268-
278.
-, I. G. Baxter and D. W. McKay. 1974. Relations between
egg production, larval production and spawning stock size in Clyde
herring, p. 1 29-1 38. In: The early life history offish. J. H. S. Blaxter
(ed.). Springer-Verlag, Berlin.
Sawada, Y. 1982. Phylogeny and zoogeography of the superfamily
Cobitoidea (Cyprinoidei, Cypriniformes). Mem. Fac. Fish. Hok-
kaido Univ. 28(2):65-223.
Sawyer, P. J. 1967. Intertidal life-history of the rock gunnel, f/!ofa
gunnellus in the western Atlantic. Copeia 1967:55-61.
Sazima, I., AND V. S. Uieda. 1979. Adaptacoes defensivas em jovens
de Oligoplites palometa (Pisces, Carangidae). Rev. Bras. Biol. 39:
687-694.
ScHEEL, J. J. 1961. Nigerian lampeye. Aquarium J. 32(9):41 1-416.
. 1962. Factors governing development of killie eggs. Tropicals
7(2):9-15.
. 1968. Rivuhns of the Old World. T.F.H. Publ. Neptune City,
ScHiNDLER, O. 1932. Sexually mature larval Hemirhamphidae from
the Hawaiian Islands. Bull. Bemice P. Bishop Mus. 197:1-28.
ScHLESiNGER, G. 1909. Zur Phylogenie und Ethologie der Scombre-
sociden. Verh. Zool. Bot. Ges. Wien 59:302-339.
Schlotterbeck, R. E., and D. W. Connally. 1982. Vertical strati-
fication of three nearshore Southern California larval fishes (En-
graulis mordax. Genyonemus linealus. and Seriphus politus). Fish.
Bull. U.S., 80:895-902.
Schmidt, J. 1904. On the pelagic post-larval halibut {Hippoglossus
vulgaris Flem. and H. hippoglossoides (Walb). Medd. Komm. Ha-
vunders. Ser. Fiskeri 1(3).
. 1905a. The pelagic post-larval stages of the Atlantic species
ofGadus. Part I. Medd. Komm. Dan. Havunders. Ser. Fiskeri 1(4).
. 1905b. On the larval and post-larval stages of the torsk (Broi-
mius hrosme (,Ascan.). Medd. Komm. Havunders. Ser. Fiskeri 1(8).
. 1 906a. The pelagic post-larval stages of the Atlantic species
of Gadus. Part II. Medd. Komm. Havunders. Ser. Fiskeri 2(2).
. 1 906b. The pelagic post-larval stages of the ling (Moha moha
(Linne) and Moha byrelange (Walbaum). Medd. Komm. Havun-
ders. Ser. Fiskeri 2(3).
. 1906c. On the larval and post-larval development of the ar-
gentines Argentina silus (Ascan.) and Argentina sphyraena (Linne)
with some notes on Maliotus villosus (O. F. Muller). Medd. Komm.
Havunders. Ser. Fiskeri 2(4).
. 1907a. On the post-larval development of the hake (Afer/uc-
cii4S vulgaris Flem.). Medd. Komm. Havunders. Ser. Fiskeri 2(7).
. 1907b. On the post-larval development of some north Atlantic
gadoids (Raniceps raninus Linne) and Moha elongata (Risso). Medd.
Komm. Havunders. Ser. Fiskeri 2(8).
. 1908. On the post-larval stages of the John Dory (Zei«ya/)er
L.) and some other Acanthopterygian fishes. Medd. Komm. Ha-
vunders. Ser. Fiskeri 2(9): 1-1 2.
— . 1913. On the identification of muraenoid larvae in their eariy
("pre-leptocephaline") stages. Medd. Komm. Havunders. Ser. Fis-
keri, 4(2): 1-1 3.
— . 1915. Respiratory adaptations of pleuronectids. Bull. Acad.
Imp. Sci. Petrograd 1915:421^44.
— . 1917. Racial investigations. I. Zoarces viviparus L. and local
races of same. C.-R. Trav. Lab. Carisberg 13(3):279-396.
— . 1918. Argentinidae, Microstomidae, Opisthoproctridae. Med-
iterranean Odontostomidae. Rep. Dan. Oceanogr. Exped. Mediterr.
2(A.5).
1921. Contributions to the knowledge of the young of the sun-
N.J.
fishes (Mola and Ranzania). Medd. Komm. Havunders. Ser. Fis-
keri 6(6):3-13.
Schmidt, P. Y. 1950. Fishes of the Sea of Okhtsk. Tr. Zool. Inst.
Akad. Nauk SSSR 6: 1-392. (Translated by Isr. Program Sci. Transl.
No. 1263).
ScHNACK, D., AND H. Rosenthal. 1978. Shrinkage of Pacific herring
larvae due to formalin fixation and preservation. Ber. Dtsch. Wiss.
Komm. Meeresforsch. 26:222-226.
Schnakenbeck, W. 1928. Beitrag zur Kenntnis der Entwicklung ei-
niger Meeresfische. 1. Ber. Dtsch. Wiss. Komm. Meeresforsch., N.F.
4(4): 119-229.
. 1931. Carangidae. Rep. Dan. Oceanogr. Exped. Mediterr.
2(A.14):l-20.
Schonevelde, S. 1624. Ichthyologia et nomenclaturae animalium
marinorum, quae in florentissimus ducatibus. Slesvici et Holstatiae,
Hamburg.
ScHULTZ, L. P. 1945. Emmelichthyops atlanticus. a new genus and
species of fish (family Emmelichthyidae) from the Bahamas, with
a key to related genera. J. Wash. Acad. Sci. 35:132-136.
. 1 946. A revision of the genera of mullets, fishes of the family
Mugilidae. with descriptions of three new genera. Proc. U.S. Natl.
Mus. 96(3204):377-395.
. 1953. Sphyraenidae. /«. Fishes of the Marshall and Marianas
Islands. Vol. 1 . Families from Asymmetrontidae through Siganidae.
Schultz, L. P., E. S. Herald, E. A. Lachner, A. D. Welander and L.
P. Woods (eds.). U. S. Natl. Mus. Bull. 202.
LITERATURE CITED
735
— . 1957. The frogfishes of the family Antennariidae. Proc. U.S.
Natl. Mus. 107:47-105.
— . 1958. Review of the parrotfishes family Scaridae. U.S. Natl.
Mus. Bull. 214.
1967. A new genus and species of zoarcid fish from the North
Pacific Ocean. Proc. U.S. Natl. Mus. 122:1-5.
ScHWARZHANS, W. 1978. Otolith-morphology and its usagc for higher
systematic units, with special reference to the Myctophiformes.
Meded. Werkgr. Ten. Kwart. Geol. 15(4): 167-1 85.
Scott, A. P., and C. Middleton. 1979. Unicellular algae as a food
for turbot (Scophthalmus maximus L.) larvae— the importance of
dietary long-chain polyunsaturated fatty acids. Aquaculture 18:227-
240.
Scott, E. O. G. 1955. Observations on some Tasmanian fishes, part
7. Pap. Proc. R. Soc. Tasmania 89:131-146.
. 1966. Observations on some Tasmanian fishes, part 14. Pap.
Proc. R. Soc. Tasmania 100:93-1 15.
. 1967. Observations on some Tasmanian fishes, part 15. Pap.
Proc. R. Soc. Tasmania 101:189-220.
Scott, J. S. 1972. Eggs and larvae of northern sand lance (Ammodytes
dubius) from the Scotian Shelf J. Fish. Res. Board Can. 29:1667-
1671.
. 1982. Distribution of juvenile haddock around Sable Island
on the Scotian Shelf J. Northwest Atl. Fish. Sci. 3:87-90.
Scott, T. D. 1962. The marine and freshwater fishes of South Aus-
tralia. W. L. Harves, Gov. Printer, Adelaide.
, C. J. M. Glover and R. V. Southcott. 1974. The marine
and freshwater fishes of South Australia. 2nd ed. A. B. James, Gov.
Printer, South Australia.
, , AND . 1980. The marine and freshwater fishes
of South Australia. 3rd ed. Gov. Printer, South Australia.
Scott, W. B., and E. J. Grossman. 1973. Freshwater fishes of Canada.
Bull. Fish. Res. Board Can. 184.
Scotten, L. M., R. E. Smith, N. S. Smith, K. S. Price and D. P.
DE SvLVA. 1973. Pictorial guide to fish larvae of Delaware Bay.
Univ. Del., Del. Bay Rep. Ser. 7.
Scrimshaw, N. S. 1945. Embryonic development in poeciliid fishes.
Biol. Bull. (Woods Hole) 88(3):233-246.
. 1946. Egg size in poeciliid fishes. Copeia 1946:20-23.
Scura, E. D., and C. W. Jerde. 1977. Various species of phytoplank-
ton as food for larval northern anchovy, Engraulis mordax. and
relative nutritional value of the dinoflagellates Gymnodinium
splendens and Gonyaulax polyedra. Fish. Bull. U.S. 75:577-583.
Seaton, D. D., and R. S. Bailev. 1971. The identification and de-
velopment of the eggs and larvae of the blue whiting Micromesislhis
poulassou (Risso). J. Cons. Int. Explor. Mer 34:76-83.
Senta. T. 1958. Young of Daicocus pelerseni i^yslTom). In: Studies
on the eggs, larvae and juveniles of Japanese fishes. Ser. 1. Uchida
et al. (eds.). Fac. Agric. Kyushu Univ. Fukuoka, Japan. (In Jap-
anese.)
. 1965. The buoyancy of sandeel eggs and their distribution in
theSeto Inland Sea. Bull. Jpn. Soc. Sci. Fish. 31:511-517.
, and S. KuMAOAi. 1977. Variation in the vertebral number of
milkfish Chanos chanos. collected from various localities. Bull. Fac.
Fish. Nagasaki Univ. 43:35-40.
Serebryakov, V. P. 1978. Development of the spotted hake, Uro-
phycis regius. from the Northwestern Atlantic. J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 18:793-799.
. 1 980. Fish eggs and larvae from the Flemish Cap area. NAFO
SCR Doc, No. 90, Senal No. N145.
Seshappa, G., AND B. S. Bhimachar. 1955. Studies on the fishery and
biology of the Malabar sole, Cvnoglossus semifasciatus Day. Indian
J. Fish. 2:180-230.
Sette, O. E. 1943. Biology of the Atlantic mackerel (5co/nfter jcom-
brns) of North Amenca: Part I. Early life history, including growth,
drift and mortality of the egg and larval populations. U.S. Fish
Wildl. Serv. Fish. Bull. 50:149-237.
, AND E. H. Ahlstrom. 1948. Estimation of abundance of the
eggs of the Pacific pilchard Sardinops caerulea off southern Cali-
fornia during 1940 and 1941. J. Mar. Res. 7:511-542.
Sha, X., AND H. RuAN. 1981. The habits and morphological devel-
opment of early stages of IHsha elongata (Bennett). Trans. Chin.
Ichthyol. Soc. 2:81-88.
, , AND G. He. 1 98 1 . The development of the egg and
larval stages of the lumpfish, Immicusjaponicus(C. and V.). Ocean-
ol. Limnol. Sin. 12(4):365-371.
Shakxee, J. B., AND C. S. Tamaru. 1981. Biochemical and morpho-
logical evolution of Hawaiian bonefishes (Albula). Syst. Zool. 30:
125-146.
, AND G. S. Whitt. 1981. Lactate dehydrogenase isozymes of
gadiform fishes: divergent patterns of gene expression indicate a
heterogeneous taxon. Copeia 1981:563-578.
Shaw, E. S. 1955. The embryology of the sergeant major, /46w/e/i/«/
saxatilis. Copeia 1955:85-99.
Shelbourne, J. E. 1964. The artificial propagation of marine fish.
Adv. Mar. Biol. 2:1-83.
. 1975. Marine fish cultivation: pioneering studies on the culture
of the larvae of the plaice (Pleuronectes platessa L.) and the sole
{Solea solea L.). Fish. Invest. Ser. II Mar. Fish. G. B. Minist. Agric.
Fish. Food, Vol. 9.
Shelton, W. L. 1978. Fate of the follicular epithelium in Dorosoma
petenense {Pisct%: Clupeidae). Copeia 1978:237-244.
, AND R. Stephens. 1980. Comparative embryogeny and early
development of threadfin and gizzard shad. Prog. Fish-Cult. 42:
34-41.
Shen, S. C. 1971. A new genus of clinid fishes from the Indo-West
Pacific, with a redescription of Clinus nematopterus. Copeia 1971:
697-706.
Shiganova, T. A. 1974. Postembryonic development oi Hygophum
benoili (Myctophidae, Pisces). J. Ichthyol. (Engl. Transl. Vopr. Ikh-
tiol.) 14:746-754.
. 1975. Postembryonic development of //v^o/i/jKW macroc/!(>
(Myctophidae, Pisces). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 15:
429^37.
. 1977. Larvae and juveniles of the lantemfishes (Myctophidae,
Pisces) of the Atlantic Ocean. Tr. Inst. Okeanol. Akad. Nauk SSSR
109:42-12.
Shiooaki, M. 1980. Redescription of the stichaeid fish C/»>o/op>i/i
saitone. Jpn. J. Ichthyol. 28:129-134.
. 1981. Life history of the stichaeid fish, Opisthocentrus ocel-
lalus. Jpn. J. Ichthyol. 29:77-85.
. 1 982. Larvae, juveniles, and a gravid female of the blennioid
Cryptacanthodes bergi, collected from Aomori, Japan. Jpn. J. Ich-
thyol. 29:102-104.
. 1983. On the life history of the stichaeid fish Chirolophis
japonicus. Jpn. J. Ichthyol. 29:446-455.
, AND Y. DoTsu. 1971a. The hfe history of the clingfish, -4.S-
pasma minima. Jpn. J. Ichthyol. 18:76-84.
, AND . 1971b. Larvae and juveniles of the clingfishes.
Lepadichlhys frenatus and Aspasmichlhys ciconiae. Jpn. J. Ichthyol.
18:85-89.
— , AND . 1971c. Egg development and hatched larva of
the clingfish, Lepadichlhys frenatus. Bull. Fac. Fish. Nagasaki Univ.
32:1-5.
— , AND . 197 Id. The life history of the clingfish, Con/rfe/w
laticephalus. Bull. Fac. Fish. Nagasaki Univ. 32:7-16.
, AND . I971e. The life history of the gobiid fish, f.vperf/o
parvulus. Bull. Fac. Fish. Nagasaki Univ. 32:17-25.
— , AND . 1972a. Life history of the blennioid fish, Dictyo-
soma burgeri. Bull. Fac. Fish. Nagasaki Univ. 33:21-38. (In Jap-
anese.)
— , AND . 1 972b. The life history of the blenniid fish Neo-
clinns byrope. Bull. Fac. Fish. Nagasaki Univ. 34:1-8. (In Japanese.)
, AND . 1 972c. The life history of the gobiid fish, Lucio-
gobius elongalus. Bull. Fac. Fish. Nagasaki Univ. 34:9-18.
— , AND . I972d. The spawning and the larva rearing of
the clingfish, Aspasma ciconiae and the spawning of the clingfish,
Aspasma minima. Bull. Fac. Fish. Nagasaki Univ. 34:39-33.
— , AND . 1973. The egg development and larva rearing of
the tripterygiid blenny, Trlplerygion etheosloma. Jpn. J. Ichthyol.
20:42-46. (In Japanese, English summary.)
736
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Shiokawa, T., AND H. TsuKAHARA. 1961. Studies On habits ofcoastal
fishes in the Amakusa Islands. I. Early life history of the purple
rockfish, Sebastes pachycephalus pachycephalus Temminck et
Schlegel. Rec. Oceanogr. Works Jpn., Spec. No. 5:123-127.
Shojima, Y. 1962. On the postlarvae and juveniles of carangid fishes
collected together with jelly-fishes. Bull. Seikai Reg. Fish. Res. Lab.
27:49-58.
, S. FujiTA AND K. UcHiDA. 1957. On the egg development and
prelarval stages of a kind of barracuda, Sphyraena pinguis Gunther.
Sci. Bull. Fac. Agric. Kyushu Univ. 16:313-318.
Shores, D. L. 1969. Postlarval Sudis (Pisces: Paralepididae) in the
Atlantic Ocean. Breviora 334.
Shuleikjn, V. V. 1958. How the pilot fish moves with the speed of
the shark. Dokl. Akad. Nauk SSSR 1 19(106): 140-143. (Transl. by
the Amer. Inst, of Biol. Sci.).
Shust, K. V. 1978. On the distribution and biology of members of
the genus Micromesislius (family Gadidae). J. Ichthyol. (Engl. Transl.
Vopr. Ikhtiol.) 18:490^*93.
SiBLEV, C. G., AND J. E. Ahlquist. 1981. The phylogeny and rela-
tionships of the ratite birds as indicated by DNA-DNA hybridiza-
tion, p. 301-335. In: Evolution today. G. G. E. Scudder and J. J.
Reveal (eds.). Proc. Sec. Int. Congr. Syst. Evol. Biol.
Silverman, M.J. 1975. Scale development in the bluefish.Po wa/owM.s
saltatrix. Trans. Am. Fish. Soc. 104:773-774.
Simons, E. v., AND J. R. VAN Horn. 1971. A new procedure for whole-
mount alcian blue staining of the cartilaginous skeleton of chicken
embryos, adapted to the clearing procedure in potassium hydroxide.
Acta Morphol. Need. Scand. 8(1970/1971):281-292.
Simonsen, M. 1978. The use of discriminant analysis in the identi-
fication of two species of larval smelt. Spinnchus ihaleichthys, and
Hypomesus I. transpacificus in the Sacramento-San Joaquin estu-
ary. Unpubl. MS thesis. Univ. of Pacific, Stockton, CA.
Simpson, A. C. 1959. The spawning of plaice in the North Sea. Fish.
Invest. Ser. II Mar. Fish. Gr. Brit. Minist. Agric. Fish. Food 22(7).
Simpson, J. G. 1959. Identification of the egg. early life history and
spawning areas of the anchoveta, Cetengraulis mysticetus (Gun-
ther), in the Gulf of Panama. Inter-Amer. Trop. Tuna Comm. Bull.
3:437-580.
Singer, C. 1950. A history of biology. Henry Schuman, New York.
. 1959. A short history of scientific ideas to 1900. Clarendon
Press, Oxford.
Slipp, J. W., and A. C. DeLacy. 1952. On the distribution and habits
of the wattled eelpout, Lycodes palearis. Copeia 1952:201-203.
Smigielskj, A. 1975a. Hormone-induced spawnings of the summer
flounder and rearing of the larvae in the laboratory. Prog. Fish-
Cuh. 37:3-8.
. 1975b. Hormonal-induced ovulation of the winter flounder,
Pseudopleuronectes amencanus. Fish. Bull. U.S. 73:431-438.
Smith, C. L. 1957. Two new clinid blennies (Atalacoclenus) from
Pueno Rico. Univ. Mich. Mus. Zool. Occas. Pap. 585:1-15.
. 1971. A revision of the American groupers: ii^j/Vifp/je/Mi and
allied genera. Bull. Am. Mus. Nat. Hist. 146:67-241.
, and R. M. Bailey. 1961. Evolution of the dorsal-fin supports
of percoid fishes. Pap. Mich. Acad. Sci. Arts Lett. 46:345-361.
, J. C. Tyler AND M. N. Feinberg. 1981. Population ecology
and biology of the pearlfish (Carapus bermudensis) in the lagoon
at Bimini, Bahamas. Bull. Mar. Sci. 31:876-902.
Smith, D. G. 1969. Xenocongrid eel larvae in the western North
Atlantic. Bull. Mar. Sci. 19:377-408.
. 1970. Notacanthiform leptocephali in the western North At-
lantic. Copeia 1970:1-9.
. 1971. Osteology and relationships of the congrid eels of the
western North Atlantic (Pisces, Anguilliformes). Unpubl. PhD dis-
sertation. Univ. of Miami, Coral Gables, FL.
. 1979. Guide to the leptocephali (Elopiformes, Anguilliformes
and Notacanthiformes). NOAA Tech. Rep. NMFS Circ. 424.
. 1980. Early larvae of the tarpon, Megalops atlantica Valen-
ciennes (Pisces: Elopidae), with notes on spawning in the Gulf of
Mexico and the Yucatan Channel. Bull. Mar. Sci. 30:136-141.
, and P. H. J. Castle. 1972. The eel genus A'eocon^er Girard:
Systematics, osteology, and life history. Bull. Mar. Sci. 22: 196-249.
, AND . 1982. Larvae of the nettastomatid eels: system-
atics and distribution. Dana-Rep. Carlsberg Found. 90:1-44.
, AND M. M. Leiby. 1980. The larva of the congrid eel Acro-
mycler alcocki {Pisces: Anguilliformes), and the distinction between
congrid and ophichthid larvae. Proc. Biol. Soc. Wash. 93(2):388-
394.
, and T. Potthoff. 1975. Leptocephali taxonomy, p. 19-20.
In: Ichthyoplankton Rep. CICAR Ichthyoplankton Workshop,
Mexico City, 17-26 July 1974. UNESCO Tech. Pap. Mar. Sci. 20.
Smith, H. M. 1907. The fishes of North Carolina. N.C. Geol. Econ.
Surv. 2:1-453.
Smith, J. L. B. 1933a. An interesting new myctophid fish from South
Africa. Trans. R. Soc. S. Afr., 21(2):I25-127.
. 1933b. The South African species of the genus //e«i;rAa«?p/!Mj
Cuv. Trans. R. Soc. S. Afr. 2 1(2): 129-1 50.
. 1935. The fishes of the family Mugilidae in South Africa. Ann.
S. Afr. Mus. 30:587-644.
. 1938. An interesting early post-larval stage of the "Galjoen."
Trans. R. Soc. S. Afr. 25:389-390.
. 1947. A generic revision of the mugilid fishes of South Africa.
Ann. Mag. Nat. Hist., Ser. II, 14:833-843.
— . 1949. The sea fishes of southern Africa. Central News Agency,
Ltd., Johannesburg.
— . 1956a. A new dealfish from South Africa. Ann. Mag. Nat.
Hist. 12:449^53.
— . 1956b. The fishes of the family Sphyraenidae in the western
Indian Ocean. Rhodes Univ. Dep. Ichthyol. Bull. 3:37-46.
— . 1958a. The gunnellichthid fishes with description of two new
species from East Africa and of Gunnellichthys (Ctarkichthys) bi-
Imeatus (C\2ir^), 1936. Rhodes Univ. Dep. Ichthyol. Bull. 9:123-
129.
— . 1958b. Fishes of the family Tetrarogidae, Caracanthidae, and
Synanciidae from the Western Indian Ocean with further notes on
scorpaenid fishes. Rhodes Univ. Dep. Ichthyol. Bull. 12:167-181.
— . 1961. The sea fishes of southern Africa. Central News Agency,
Ltd., South Africa.
— . 1963. Fishes of the families Draconettidae and Callionymidae
from the Red Sea and the western Indian Ocean. Rhodes Univ.
Dep. Ichthyol. Bull. 28:547-564.
— . 1966a. Fishes of the sub-family Nasinae with a synopsis of
the Prionurinae. Rhodes Univ. Dep. Ichthyol. Bull. 32:635-681.
. 1966b. Hitherto unknown early developmental larval stadia
of the west African albulid fish, Pterothrissus belloci Cadenat, 1937.
Rhodes Univ. Dep. Ichthyol. Occas. Pap. (6):57-61.
. 1968. A new liparine fish from the Red Sea. J. Nat. Hist. 2:
105-109.
Smith, P. E. 1972. The increase in spawning biomass of northern
anchovy Engraults mordax. Fish. Bull. U.S. 70:849-874.
, and S. L. Richardson. 1977. Standard techniques for pelagic
fish egg and larval surveys. FAO Fish. Tech. Pap. No. 175.
, AND . 1979. Selected bibliography on pelagic fish egg
and larva surveys. FAO Fish. Circ. No. 706.
Smith, S. 1957. Early development and hatching, p. 323-359. /«.■ The
Physiology of Fishes. Vol. 1. M. E. Brown (ed). Academic Press,
New York.
Smith, W. G., AND M. P. Fahay. 1970. Description of eggs and larvae
of the summer flounder, Paralichthys dentatus. U.S. Fish Wildl.
Serv. Res. Rep. 75.
Smith-Vaniz, W. F. 1975. Supplemental description of a rare blenniid
fish Phenablennius heyligen (Bleeker). Proc. Acad. Nat. Sci. Phila.
l27(6):53-55.
. 1976. The saber-toothed blennies, tribe Nemophini (Pisces:
Blenniidae). Proc. Acad. Nat. Sci. Phila. Monogr. 19:1-196.
, AND F. J. Palacio. 1974. Atlantic fishes of the genus Acan-
themblemaria. with descriptions of three new species and com-
ments on Pacific species (Clinidae: Chaenopsinae). Proc. Acad. Nat.
Sci. Phila. 125( 1 l):197-224.
LITERATURE CITED
737
— , AND V. G. Springer. 1971. Synopsis of the tribe Salariini,
with description of five new genera and three new species (Pisces:
Blenniidae). Smithson. Contrib. Zool. 73:1-72.
-, AND J. C. Staiger. 1973. Comparative revision oi Scomber-
oides, Oligoplttes. Parana, and Hypacanthus with comments on the
phylogenetic position of Campo^ramma (Pisces: Carangidae). Proc.
Calif. Acad. Sci. 39:185-256.
Smitt, F. A. 1895. A history of Scandinavian fishes, II. Stockholm
andLondon:567-1240.
Smoker, W., AND W. C. Pearcv. 1970. Growth and reproduction of
the lantemfish Stenobrachius lecopsarus. J. Fish. Res. Board Can.
27:1265-1275.
Smol'Yanov, I. I. 1961. The development of Brachymyslax lenok
(Pallas). Vopr. Ikhtiol. 18:136-148. (In Russian.)
Sneath, P., and R. SoKAL. 1973. Numerical taxonomy. W. H. Free-
man, San Francisco.
Snyder, D. E. 1976. Terminologies for intervals of larval fish devel-
opment, p. 4 1-58. In: Great Lakes fish egg and larvae identification:
Proceedings of a workshop. J. Boreman(ed.). U.S. FishWildl. Serv.
Biol. Serv. Prog. FWS/OBS-76/23.
. 1979. Burbot: larval evidence for more than one North Amer-
ican species, p. 204-219. In: Proc. 3rd Symp. Larval Fish. West.
Kentucky Univ. Dep. Biol., Bowling Green.
. 1981. Contributions to a guide to the cypriniform fish larvae
of the upper Colorado River system. U.S. Dep. Inter.. Bur. Land
Manage., Colo. Off., Biol. Sci. Sen 3.
, AND S. C. Douglas. 1978. Description and identification of
mooneye, Hiodon tergtsus. protolarvae. Trans. Am. Fish. Soc. 107:
590-594.
, M. B. M. Snyder and S. C. Douglas. 1977. Identification
of golden shiner, Notemigonus crysoleucas, spotfin shiner, Nolropis
spilopterus. and fathead minnow, Pimephales promelas. larvae. J.
Fish. Res. Board Can. 34:1397-1409.
Soin, S. C. 1966. Development types of structures and phylogeny of
vascular systems of the vitelline sac in fish embryos, performing a
respiratory function. Zool. Zh. 45:1382-1397. (In Russian.)
. 1968. Some features in the development of the blenny[ZoarcM
viviparus (L.)] in relation to viviparity. Prob. Ichthyol. (Eng. Transl.
Vopr. Ikhtiol.) 8:222-229.
1980. Types of development of salmoniform fishes and their
taxonomic importance. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.)
20(l):49-56.
Sokal, R., AND Sneath, P. 1963. Principles of numerical taxonomy.
W. H. Freeman, San Francisco.
SoLDATOv, V. K., and G. U. LiNDBERG. 1930. A review of the fishes
of the seas of the Far East. Izv. Tikhookean. Nauchno-Issled. Inst.
Rybn. Khoz. Okeanogr. 5.
SOLTZ, D. L., AND M. F. Hirschreld. 1981. Genetic differentiation
of pupfishes (Genus Cyprinodon) in the American southwest, p.
291-333. In: Fishes in North American Deserts. R. J. Naiman and
D. L. Soltz (eds.). John Wiley & Sons, New York.
SooNG, M. H. H. 1968. Aspects of reproduction in freshwater half-
beaks. Malay. Nat. J. 21(suppl.):xxxiii-xxxiv.
Sparta, A. 1926. Uova e larve di Fierasfer dentatus Cuv. Mem. R.
Com. Talassogr. Ital. 122.
. 1 928. Stadi larvali di Uraleptus maraldi Risso. Mem. R. Com.
Talassogr. Ital. 141.
. 1932. Contributo alia conoscenza dello sviluppo post-em-
brionale in Anthias sacer Bloch. Mem. R. Com. Talassogr. Ital.
195.
. 1933. Ordine: Allotriognathi, p. 266-279. In: Uova. larve e
stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1934. Uova e larve di Gobiidae. I. GoW«ipaga/!c//u5 L. Mem.
R. Com. Talassogr. Ital. 211.
. 1935. Contributo alia conoscenze dello sviluppo nei Percidi.
Uova ovariche mature di Epinephelus giia:a L. e stadi post-em-
brionali e larvali di Epinephelus alexandrinus Cuv. e Val. Mem.
R. Com. Talassogr. Ital. 224.
. 1936. Contributo alia conoscenza di uova, stadi embrionali e
post-embrionali in Macrorhamphosus scolopax L. Mem. R. Com.
Talassogr. Ital. 225.
— . 1937. Contributo alia conoscenza dello sviluppo embrionale
e post-embnonale nei Murenoidi. \. — Coecu!a imberbis (De la
Roche). 2. — Ophichthys remicaudus (Kaup). Mem. R. Com. Tal-
assogr. Ital. 244.
— . 1938a. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. i. — Coecula coeca (L.). Mem. R.
Com. Talassogr. Ital. 248.
— . 1938b. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. 4. — Ophisoma balearicum De la
Roche. Mem. R. Com. Talassogr. Ital. 252.
— . 1939a. Contributo alia conoscenza dello sviluppo nei Percidi.
Uova, stadi embrionali e post-embrionali di Polyprion cernium
(Val.) Mem. R. Com. Talassogr. Ital. 259.
— . 1939b. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. 5. — Nuova specie di uovo di
Murenoidi (Conger congerl). Mem. R. Com. Talassogr. Ital. 264.
— . 1 939c. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. (>. — Ophisurus serpens L. Mem.
R. Com. Talassogr. Ital. 267.
— . 1 939d. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. 1 . — Chlopsis bicolor('R.ai.). Mem.
R. Com. Talassogr. Ital. 268.
— . 1 940a. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. %. — Echelus myrus L. Mem. R.
Com. Talassogr. Ital. 273.
— . 1 940b. Contributo alia conoscenza dello sviluppo embrionale
e post-embrionale nei Murenoidi. 9. — Seconda nuova specie di uova
di Murenoidi. Mem. R. Com. Talassogr. Ital. 279.
— . 1 942a. Uova, stadi embrionali, prelarve e larve di Nettastoma
melanurum Raf. Arch. Oceanogr. Limnol. 2:151-158.
. 1942b. Uova e larvae di Aiherina rissoi C. V. ottenute da
fecondiazione artificiale, e stadi post-embrionali raccolti nei planc-
ton. Arch. Oceanogr. Limnol. 2:161-168.
. 1946. Contributo alia conoscenza dello sviluppo post-em-
brionale in Cubiceps gracilis Lowe. Boll. Pesca Piscic. Idrobiol.
22(1).
. 1948. Uova ovariche, uova fecondate tenute in coltura, larva
alia schiusa, stadi larvali e giovanili di Cristiceps argentalus Risso.
Arch. Oceanogr. Liminol., Mem. 315, 5(l-3):79-84.
. 1951. Rarissimi stadi larvali e giovanili di Myctophum glaciale
(Reinhardt). Boll. Pesca Piscic. Idrobiol. 6(2):101-107.
. 1952. Contributo alia conoscenza dello sviluppo larvale di
Myctophum melopoclampum Cocco. Boll. Pesca Piscic. Idrobiol.
7(i):5-10.
. 1953. Uova e larve di Tetrapturus belone Raf (Aguglia Im-
perial). Boll. Pesca Piscic. Idrobiol. 8(l):58-62.
. 1954. Desstadesjeunesraresde teleosteens marins((7ra/ep/M.s
maraldii Risso, Temnodon saltalor C. V., Lophotes cepedianus
Gioma). Rapp. P. V. Reun. Comm. Int. Explor. Sci. Mer. Mediterr.
Monaco 12:79-84.
— . 1956a. Sottordine: Pharyngognathi J. Miiller, p. 576-598. In:
Uova, larve e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli
Monogr. 38.
. 1956b. Sottordine: Scleroparei. Famiglia 1: Scorpaenidae, p.
599-626. In: Uova. larve e stadi giovanili de Teleostei. Fauna Flora
Golfo Napoli Monogr. 38.
— . 1962. Uova, larve alia schiusa ed al 4 giomo de vita, Po-
matomus sallalrix GiW. (Syn. Temnodor sallator. C.V.). Boll. Pesca
Piscic. Idrobiol. l7(l):5-8.
Sponaugle, D. L, and J. P. Wourms. 1979. Secondary egg envelope
pre-patteming by follicle cells during fish oogenesis. Am. Zool. 19:
946.
Springer, V. G. 1954. Western Atlantic fishes of the genus /"aracZ/wMi:.
Tex. J. Sci. 4:422-441.
. 1955. The taxonomic status of the fishes of the genus 5;ar/i-
monotus. including a review of the Atlantic species. Bull. Mar. Sci.
Gulf Caribb. 5:66-80.
738
ONTOGENY AND SYSTEM ATICS OF FISHES -AHLSTROM SYMPOSIUM
— . 1958. Systematics and zoogeography of the clinid fish of the
subtribe Labrisomini. Tex. Inst. Mar. Sci., Hubbs Publ. 5:4 1 8-492.
— . 1959. A new species of LaZimowiis from the Caribbean Sea,
with notes on other fishes of the subtribe Labrisomini. Copeia 1959:
289-292.
— . 1967. Revision of the circumtropical shorefish genus ifmo-
macro^^iis (Blenniidae: Salariinae). Proc. U.S. Natl. Mus. 1 22(3582):
1-150.
— . 1968. Osteology and classification of the fishes of the family
Blenniidae. U.S. Natl. Mus. Bull. 284.
— . 1970. The western south Atlantic clinid fish Ribeiroclinus
eigenmanni. with discussion of the intrarelationships and zooge-
ography of the Clinidae. Copeia 1970:430-436.
— . 1971. Revision of the fish genus Ecsenius (Blenniidae. Blen-
niinae, Salariini). Smithson. Contrib. Zool. 72.
— . 1972a. Synopsis of the tribe of Omobranchini with descrip-
tions of three new genera and two new species (Pisces: Blenniidae).
Smithson. Contrib. Zool. 130.
— . 1972b. Additions to revisions of the blenniid fish genera £c-
senius and Enlomacrodus, with descriptions of three new species
of Ecsenius. Smithson. Contrib. Zool. 134.
— . 1976. C;>ma/ur;i« iMwarei. new genus and species of blenniid
fish from the Indian Ocean. Proc. Biol. Soc. Wash. 89(3): 199-203.
— . 1978. Synonymization of the family Oxudercidae, with com-
ments on the identity of Apocryptes cantons Day (Pisces: Gobiidae).
Smithson. Contrib. Zool. 270.
— . 1982. Pacific plate biogeography, with special reference to
shorefishes. Smithson. Contrib. Zool. 367.
— . 1983. Tyson belos, new genus and species of western Pacific
fish (Gobiidae, Xenisthminae), with discussions of gobioid osteol-
ogy and classification. Smithson. Contrib. Zool. 390.
— . MS. A new genus and species of western Pacific fish (Gobiidae,
Xenisthminae), with discussions of gobioid osteology and classi-
fication.
— , AND T. H. Eraser. 1976. Synonymy of the fish families Chei-
lobranchidae (=Alabetidae) and Gobiesocidae, with descriptions of
two new species of Alabes. Smithson. Contrib. Zool. 234.
— , AND M. F. GoMON. 1975a. Revision of the blenniid fish genus
Omobranchus with descriptions of three new species and notes on
other species of the tribe Omobranchini. Smithson. Contrib. Zool.
177.
— , AND . 1975b. Variation in the western Atlantic clinid
fish Malacoctenus iriangulatus with a revised key to the Atlantic
species of Malacoctenus. Smithson. Contrib. Zool. 200.
, C. L. Smith and T. H. Fraser. 1977. New species of pro-
togynous hermaphroditic fish, and synonymy of Anisochromidae,
Pseudoplesiopidae, and Pseudochromidae. Smithson. Contrib. Zool.
252.
, AND W. F. Smith-Vaniz. 1972. A new iribe (Phenablennitni)
and genus (Phenablennius) of blenniid fishes based on Petroscirtes
heyligeri (Bleeker). Copeia 1972:64-71.
, AND A. E. Spreitzer. 1978. Five new species and a new genus
of Indian Ocean blenniid fishes, tribe Salariini, with a key to genera
of the tribe. Smithson. Contrib. Zool. 268.
Staioer, J. C. 1965. Atlantic flying fishes of the genus Cypselurus,
with descriptions of the juveniles. Bull. Mar. Sci. 15:672-725.
Stark, W. A., II. 1971. Biology of the gray snapper, Li^f/a/iMjgr/ieiis
(Linnaeus), in the Florida Keys, p. 1 1-150. In: Investigations on
the gray snapper, Lutjanus griseus. W. A. Stark, II, and R. E. Schroe-
der (eds.). Stud. Trop. Oceanogr. (Miami) 10.
Starks, E. C. 1 900. The osteological characters of the fishes of the
Suborder Percesoces. Proc. U.S. Natl. Mus. 22:1-10.
. 1902. The shoulder girdle and characteristic osteology of the
hemibranchiate fishes. Proc. U.S. Natl. Mus. 25:619-634.
. 1911. The osteology and relationships of fishes belonging to
the family Carangidae. Leland Stanford Jr. Univ. Publ. Univ. Ser.
5:27-49.
Steedman, H. F. 1976. General and applied data on formaldehyde
fixation and preservation of marine zooplankton, p. 103-154. In:
Zooplankton fixation and preservation. H. F. Steedman (ed.).
UNESCO Press, Paris.
Steftensen, E. 1980. Daily growth increments observed in otoliths
from juvenile East Baltic cod. Dana 1:29-37.
Stehmann, M., and W. Lenz. 1973. Ergebnisse der Forschimgsreisen
des FFS "Walther Herwig" nach Sudamerika. XXVI. Systematik
und Verbreitung der Artengruppe— 5er/o/e//a punctata (Schneider,
1801), S. porosa Guichenot, 1848, 5. dobula Gunther, 1869)-
sowie taxonomische Bemerkungen zu Hyperoglyphe Gunther, 1859
und Schedophilus Cocco, 1839 (Osteichthyes, Stromateoidei, Cen-
trolophidae). Arch. Fischereiwiss. 23:179-201.
Stehr, C. M. 1979. The development of the hexagonal surface struc-
ture of the C-O sole {Pleuronichthys coenosus). Am. Zool. 19:976.
. 1982. The development of the hexagonally structured egg
envelope of the C-O sole (Pleuronichthys coenosus). Unpubl. MS
thesis, Univ. Washington, Seattle.
, and J. W. Hawkes. 1979. The comparative ultrastructure of
the egg membrane and associated pore structures in the starry floun-
der, Platichthys stellatus (Pallas), and pink salmon, Oncorhynchus
gorbuscha (Walbaum). Cell Tissue Res. 202:347-356.
Stein, D. L. 1978. A review of the deepwater Liparidae (Pisces) from
the coast of Oregon and adjacent waters. Occ. Pap. Calif Acad.
Sci. No. 127.
. 1 980a. Aspects of reproduction of liparid fishes from the con-
tinental slope and abyssal plain off" Oregon, with notes on growth.
Copeia 1980:687-699.
. 1980b. Description and occurrence of macrourid larvae and
juveniles in the northeast Pacific Ocean off Oregon, U.S.A. Deep-
Sea Res. 27A:889-900.
, AND C. E. Bond. 1978. A new deep-sea fish from the eastern
North Pacific Psychrolutes phrictus (Pisces: Cottidae [Psychroluti-
dae]). Nat. Hist. Mus. Los Ang. Cty. Contrib. Sci. 296:1-9.
, AND W. G. Pearcy. 1982. Aspects of reproduction, early life
history, and biology of macrourid fishes off" Oregon, U.S.A. Deep-
SeaRes. 29:1313-1300.
Stein, R. 1972. Identification of some larval Pacific cottids. Unpubl.
MS Thesis. Humboldt State Coll., Areata, CA.
. 1973. Description of laboratory-reared larvae of 0%ocoHitf
macutosus GmYd (Pisces: Cottidae). Copeia 1973:373-377.
Stephens, J. S. 1963. A revised classification of the blennioid fishes
of the American family Chaenopsidae. Univ. Calif Publ. Zool. 68.
. 1970. Seven new chaenopsid fishes from the western Atlantic.
Copeia 1970:280-309.
— , E. S. Hobson and R. K. Johnson.
1 966. Notes on distribution.
behavior, and morphological variation in some chaenopsid fishes
from the tropical eastern Pacific, with descriptions of two new
species, Acanthemblemaria castroi and Coralliozetus springeri. Co-
peia 1966:424-438.
— , AND V. G. Springer. 1974. Clinid fishes of Chile and Peru,
with description of a new species, Myxodes ornattis from Chile.
Smithson. Contrib. Zool. 1 59.
Stephenson, A. B. 1977. Metavelifer mulliradiatus (Regan, 1907)
(Pisces: Veliferidae), a new record from New Zealand waters. Rec.
Auckl. Inst. Mus. 14:143-144.
Sterba, G. 1962. Die Schmerlenartigen (Cobitidae), p. 201-234. In:
Handbuch der Binnenfischerei Mitteleuropas. 3B(9). R. Demoll and
H. N. Maier (eds.). E. Schweizerbart'sche Verlagsbuchhandlung,
Stuttgart, West Germany.
, andH. MOller. 1962. Elektronenmikroskopische untersuch-
ungen iiber Bildung und Struktur der Eihiillen bei Knochenfischen,
I. Die Hiillen junger Oocyten von Cynolebias bellotttt. Steindachner
(Cyprinodontidae). Zool. Jb. Abt. Anat. 80(l):65-80.
Stevens, E. G., and H. G. Moser. 1982. Observations on the early
life history of the mussel blenny, Hypsoblennius jenkinsi. and the
bay blenny, Hypsoblennius gentilis, from specimens reared in the
laboratory. Calif. Coop. Oceanic Fish. Invest. Rep. 23:269-275.
Steyskal, G. C. 1980. The grammar of family-group names as ex-
emplified by those of fishes. Proc. Biol. Soc. Wash. 93:168-177.
Stiassny, M. L. J. 1981. The phyletic status of the family Cichlidae
LITERATURE CITED
739
(Pisces, Perciformes): A comparative anatomical investigation. Neth.
J. Zool. 31(2);275-314.
Strasburg, D. W. 1 960. A new Hawaiian engraulid fish. Pac. Sci. 14:
395-399.
. 1961. Larval carapid fishes from Hawaii, with remarks on the
ecology of adults. Copeia 1961:478-480.
. 1962. Pelagic stages of Za«c/«i canMce^w from Hawaii. Co-
peia 1962:344-345.
. 1964. Postlarval scombroid fishes of the genera ^camAocyt-
ium. Nealotus and Diplospinus from the Central Pacific Ocean. Pac.
Sci. 18:174-185.
-. 1965. Description of the larva and familial relationships of
the fish Snyderidia camna. Copeia 1965:20-24.
Stroud, G. J., B. Goldman, F. H. Talbot and W. Gladstone. MS.
Development, growth and age determination of the short-nosed
pufferfish Canthigaster valenlini (Pisces: Tetraodontidae).
Struhsaker, P., and J. H. Uchiyama. 1976. Age and growth of the
nehu Slolephorus purpureus (Pisces: Engraulidae) from the Hawai-
ian Islands as indicated by daily growth mcrements of sagittae.
Fish. Bull. U.S. 74:9-17.
Suarez, S. S. 1975. The reproductive biology of Og//Wa cayoruwj, a
viviparous brotulid fish. Bull. Mar. Sci. 25:143-173.
Sudarsan, D. 1968a. On the early development of the pipe-fish Sy«-
gnathoides biaculealus (Bloch). J. Mar. Biol. Assoc. India 8:222-
224.
Sudarsan, D. 1968b. Eggs and larvae of a hemirhamphid fish from
Mandapam. J. Mar. Biol. Assoc. India 8:342-346.
SuLAK, K. J. 1977. The systematics and biology of Sar/!yp?eroK (Pisces,
Chlorophthalmidae) with a revised classification of benthic myc-
tophiform fishes. Galathea Rep. 14:49-108.
SuLYA, L. L., B. E. Box AND G. GuNTER. 1960. Distribution of some
blood constituents in fishes from the Gulf of Mexico. Am. J. Phys-
iol. 199:1177-1180.
SuMiDA, B., E. H. Ahlstrom and H. G. Moser. 1979. Early devel-
opment of seven flatfishes of the eastern North Pacific with heavily
pigmented larvae (Pisces. Pleuronectiformes). Fish. Bull. U.S. 77:
105-145.
, AND H. G. MosER. 1980. Food and feeding of Pacific hake
larvae, Merluccius productus. off' southern California and northern
Baja California. Calif Coop. Oceanic Fish. Invest. Rep. 21:161-
171.
Sunny, K. G. 1971. Morphology of the vertebral column of Mugil
macrolepis (Smith). Bull. Dep. Mar. Biol. Oceanogr. Univ. Cochin
5:101-108.
Suzuki, K. 1962. Anatomical and taxonomical studies on the carangid
fishes of Japan. Rep. Fac. Fish. Prefect. Univ. Mie 4(2):43-232.
, AND S. HioKL 1978. Spawning behavior, eggs, and larvae of
the sea bream, Gymnocranius griseus, iiran aquarium. Jpn. J. Ich-
thyol. 24:271-277.
, AND . 1979a. Morphology and phyletic aspects of the
newly hatched larvae, with special references to those of family
Serranidae. Kaiyo Kaguku (Mar. Sci.) 11:1 17-125. (In Japanese).
, AND . 1979b. Spawning behavior, eggs, and larvae of
the lutjanid fish, Liiljanus kasmira. in an aquarium. Jpn. J. Ich-
thyol. 26:161-166.
, , K. KoBAYASHi AND Y. Tanaka. 1981. Developing
eggs and early larvae of the wxasses, Cirrhilabrus temminckii and
Labroides diinidiatus. with a note on their spawning behaviors. J.
Fac. Mar. Sci. Technol. Tokai Univ. 14:369-372. (In Japanese,
English Abstr.)
, , Y. Tanaka andK. IwASA. 1979. Spawning behaviour,
eggs, larvae, and sex reversal of two Pomacanthine fish, Genican-
ihus leinarck and G. semifascialus. in the aquarium. J. Fac. Mar.
Sci. Technol. Tokai Univ. 12:149-165.
, K. KoBAYASHi, S. HioKj AND T. SAKAMOTO. 1974. Ecological
studies of the anthiine fish Sacura margahtacea in Suruga Bay.
Jpn. J. Ichthyol. 21:21-33.
, , AND . 1978. Ecological studies of the
anthiine fish, Franzia squamipinnis in Suruga Bay, Japan. Jpn. J.
Ichthyol. 25:124-140. (In Japanese, English abstract.)
— , , AND . In press. Ecological studies of the
anthiine fish, Pseudanthias elongatus. in Suruga Bay, Japan.
— , H. Syozo, Y. Tanaka and H. Kjtazawa. 1983. Spawning
and early life history of Hapalogenys mucronatus (Eydoux et Sou-
leyet) (Pisces: Pomadasyidae) in an aquarium. J. Fac. Mar. Sci.
Technol. Tokai Univ. 16:183-191.
— , Y. Tanaka and S. Hioki. 1 980. Spawning behaviour, eggs,
and larvae of the butterflyfish, Chaetodon nippon, in an aquarium.
Jpn. J. Ichthyol. 26:334-341.
— , , AND Y. Shiobara. 1980. Studies on repro-
duction and larval rearing of coastal marine fishes, p. 53-82. In:
Research in a large-scale culture of marine fisheries resources. G.
Yamamoto (ed.). Inst. Oceanic Res. and Dev., Tokai Univ., Shim-
izu. (In Japanese.)
— , , AND S. Ueno. 1981. Spawning behavior and
early life history of Hypodytes rubripinnis (Pisces: Congiopodidae)
in an aquarium. J. Fac. Oceanogr. Tokai Univ. 14:357-367.
Suzuki, R. 1955. Note on the interfamiliar hybrid larvae between
mud loach (Misgitrnus anguitlicaudatus) and Crucian carp (Car-
assius carassius) or goldfish (Carassius auratus). Jpn. J. Ichthyol.
4:50-58.
. 1968. Hybridization experiments in cyprinid fishes. XI. Sur-
vival rate of F, hybrids with special reference to the closeness of
taxonomical position of combined fishes. Bull. Freshwater Fish.
Res. Lab. (Tokyo)18(2):l 13-155.
Sverdrup, H. U., M. W. Johnson and R. H. Fleming. 1942. The
Oceans. Their physics, chemistry, and general biology. Prentice-
Hall, Inc., Englewood Cliffs, NJ.
SvETOviDOv, A. N. 1937. Uber die Klassifikation der Gadiformes oder
Anacanthini. Bull. Acad. Sci. URSS, p. 1281-1288.
. 1948. Treskoobraznye (Gadiformes). Akad.NaukSSSR, Zool.
Inst., Fauna SSSR 9(4), N.S. 34. (In Russian, transl. by Isr. Program
Sci. Transl., 1962. Avail, from: NTIS, Springfield, VA., OTS 63-
11071.)
. 1956. Morphological principles of the classification of the
Gadidae. Proc. Int. Congr. Zool. 14(1953):535-540.
. 1963. Fauna of U.S.S.R. Fishes. Vol. 2, No. 1. Clupeidae.
Transl. from Russian by Isr. Program Sci. Transl. OTS 61-1 1435.
(First published in 1952, Zool. Inst. Akad. Nauk SSSR, N.S. 48.)
. 1969. On the taxonomical position of the genus £Mc/;c/i;/iyi
(Pisces, Gadiformes). Zool. Zh. 48:1824-1831.
-. 1982. Taxonomic differences between red hake. Urophycis
chuss, and U. tenuis (Gadidae). J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 22(3): 150-1 53.
SwARUP, H. 1958. Stages in the development of the stickleback Gar-
terosteus oculealus (L.). J. Embryol. Exp. Morphol. 6(3):373-383.
Takagi, K. 1950. On the glossohyal bone of the gobioid fishes of
Japan, with some phylogenetic considerations. Jpn. J. Ichthyol.
1:37-52. (In Japanese.)
. 1953. A study on the scale of the gobiid fishes of Japan. J.
Tokyo Univ. Fish. 39:231-253.
-. 1966. Distributions and ecology of the gobioid fishes in the
Japanese waters. J. Tokyo Univ. Fish. 52:83-127.
Takai, T., and T. Fukunaoa. 1971. The life history of a ovo- vivip-
arous scorpaenoid fish, Sebasles longispinis (Matsubara). 1 . Egg and
larval stages. J. Shimonoseki Univ. Fish. 20:91-95.
, and a. Mizokami. 1959. On the reproduction, eggs, and larvae
of the pipefish, Synagnathus schlegeli Kaup. J. Shimonoseki Coll.
Fish. 8:85-89.
AND T. YosHioKA. 1979. The life history of the dragonet.
CaUionymus beniteguri in the Seto Inland Sea. I. Egg, larva and
juvenile. J. Shimonoseki Univ. Fish. 27:147-154.
Takita, T. 1966. Egg development and larval stages of the small
clupeoid fish, Harengula -unasi Bleeker and some information about
the spawning and nursery in Ariake Sound. Bull. Fac. Fish. Na-
gasaki Univ. 21:171-179.
. 1967. The spawning and early hfe history of the engraulid fish
740
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Coilia sp. distributed in Ariake Sound. Bull. Fac. Fish. Nagasaki
Univ. 23:107-122.
— . 1980. Embryonicdevelopment and larvae of three dragonets.
Bull. Jap. See. Sci. Fish. 46:1-7.
— . 1983. Embryonic and larval development of the callionymid
fish, Callionymus calUste. Jpn. J. Ichthyol. 29:441-445.
— , AND S. FujiTA. 1964. Egg development and prolarval stages
of the turbot, Pleuronichthys cornulus (Temminck et Schlegel). Bull.
Jpn. Soc. Sci. Fish. 30:613-618. (In Japanese, English summary.)
— , andY. DoTSu. 1967. Egg development and pro-larval
stages of flatfish, Verasper variegalus (Temminck et Schlegel). Bull.
Fac. Fish. Nagasaki Univ. 23:101-106.
, AND E. Okamoto. 1979. Spawning behavior of the two dra-
gonets, Callionvmus flagris and C rkhardsoni in the aquarium.
Jpn. J. Ichthyol. 26:282-288.
Taliev, D. N. 1955. Bullheads and sculpins (Cottoidei) from Lake
Baikal. Akad. Nauk. SSSR. (In Russian.)
Talwar, p. K. 1968. Studies on the biology of Hemirhamphus mar-
ginatus (Forskal) (Hemirhamphidae-Pisces). J. Mar. Biol. Assoc.
India 9:61-69.
TAning, a. V. 1918. Mediterranean Scopelidae (Saurus aulopus.
Chlorophthalmus. and Myctophum). Rep. Dan. Oceanogr. E.xped.
Mediterr. 2(A.7).
. 1923. LopA/us. Rep. Dan. Oceanogr. Exped. Mediterr. 2( A. 10).
. 1931. PostlarvalstagesofBa//i.v/agi« from the North Atlantic.
Vidensk. Medd. Dann. Naturhist. Foren. Khobenhavn 92:269-274.
. 1936. On the eggs and young stages of the halibut. Medd.
Komm. Havunders. Ser. Fiskeri 10(4).
. 1952. Experimental study of meristic characters in fishes. Biol.
Rev. 27:169-193.
. 1955. On the breeding areas of the swordfish (Xiphias). Deep-
Sea Res. 3 (Suppl.):438-450.
. 1961. Larval and postlarval stages of Sebastes species and
Heticolenus daciylopterus. Rapp. P.V. Reun. Cons. Int. Explor. Mer
150:234-240.
Taranets, S. Y. 1941. On the classification and ongm of the family
Cottidae. Izv. Akad. Nauk SSSR Otd. Biol. 194l(3):427^47.
(Translated from Russian, Univ. Brit. Columbia Mus. Contrib. 5.)
Taranetz, A. Y. 1935. Briefsurvey of species of the genus forofo»M.s.
Bull. Far Eastern Branch Acad. Sci. USSR 14:177-180.
. 1937. Handbook for identification of fishes of Soviet, Far East
and adjacent waters. Bull. Pac. Sci. Inst. Fish. Oceanogr. 11(11):!-
200.
Tarp, F. H. 1952. A revision of the family Embiotocidae (the
surfperches). Calif Dep. Fish Game Fish Bull. 88:1-99.
Ta VOLGA, W. N. 1949. Embryonic development of the platyfish
(Platypoecilius) and swordtail (Xiphophorus) and their hybnd. Bull.
Am. Mus. Nat. Hist. 94:165-229.
Taylor, J. N. 1983. Field observations on the reproductive ecology
of three species of armored catfishes (Loricariidae: Loricariinae) in
Paraguay. Copeia 1983:257-259.
Taylor, M. N., L. DiMichele and G. J. Leach. 1977. Egg stranding
in the life cycle of the mummichog. Fundulus heterocUtus. Copeia
1977:397-399.
Taylor, W. R. 1964. Fishes of Arhnem Land. Rec. Am. Aust. Sci.
Exped. Amhem Land. 4:45-307.
. 1967. An enzyme method of clearing and staining small ver-
tebrates. Proc. U.S. Natl. Mus. 122(3596): 1-1 7.
. 1977. Observations on specimen fixation. Proc. Biol. Soc.
Wash. 90:753-763.
, AND G. C. Van Dyke. MS. Staining and clearing small ver-
tebrates for bone and cartilage study.
Templeman, W. 1 948. The life history of the capelin, Malhtus villosus
in Newfoundland waters. Res. Bull. Div. Fish. Res. Newfoundland
17:1-151.
. 1968. A review of the morid fish genus //a/a/-g3T«« with first
records from the western North Atlantic. J. Fish. Res. Board Can.
25:877-901.
Terner, C. 1979. Metabolism and energy conversion, p. 261-278. /«.
Fish Physiology. Vol. 8. W. S. Hoar, D. J. Randall and J. R. Brett
(eds.). Academic Press, New York.
Thangaraja, M., and K. Ramamoorthl 1980. Early life history of
Anadontostoma chacunda (Ham.-Buch.) from Porto Novo. Indian
J. Mar. Sci. 9:136-139.
Theilacker, G. H. 1978. Effect of starvation on the histological and
morphological characteristics of jack mackerel Trachurus sym-
metncus larvae. Fish. Bull. U.S. 76:403-414.
. 1980a. Changes in body measurements of larval northern
anchovy Engraulis mordax and other fishes due to handling and
preservation. Fish. Bull. U.S. 78:685-692.
. 1980b. Rearing container size affects morphology and nutri-
tional condition of larval jack mackerel, Trachurus svmmetricus.
Fish Bull. U.S. 78:789-791.
, AND K. Dorsey. 1980. Larval fish diversity, a summary of
laboratory and field research. IOC Workshop Rep. No. 28:105-
142.
, andM. F. McMaster. 1971. Mass culture of the rotifer Brac-/;-
ionus plicatiHs and its evaluation as a food for larval anchovies.
Mar. Biol. 10:183-188.
Thibault, R. E., and R. J. ScHiiLTZ. 1978. Reproductive adaptation
among viviparous fishes (Cyprinodontiformes: Poeciliidae). Evo-
lution 32:320-332.
Thomopoulos, A. 1954. Sur quelques oeufs planctoniques de teleos-
teens de la Baie de Villefranche. Bull. Inst. Oceanogr. (Monaco)
1043:1-15.
. 1956. Sur quelques oeufs planctoniques de teleosteens de la
Baie de Villefranche. II. Peches de mois de Septembre. Bull. Inst.
Oceanogr. Monaco 1072.
Thompson, W. 1840. On a new genus of fishes from India. Charles-
worth's Mag. Nat. Hist. 4:184-187.
Thompson, W. F. 1916. Fishes collected by the United States Bureau
of Fisheries steamer "Albatross" during 1888, between Montevi-
deo, Uruguay, and Tome, Chile, on the voyage through the Straits
of Magellan. Proc. U.S. Natl. Mus. 50(2 133):40 1^76.
, and J. B. Thompson. 1919. The spawning of the grunion {Leu-
resthes tenuis). Calif Fish Game Fish Bull. 3:1-29.
, and R. Van Cleve. 1936. Life history of the Pacific halibut.
2. Distribution and early life history. Rep. Int. Fish. Comm. 9.
, F. H. Bell, H. A. Dunlop, L. P. Schultz and R. Van Cleve.
1936. The spawning of the silver smelt Hypomesus pretiosus. Ecol-
ogy 17:158-168.
Thomson, G. M. 1906. The Portobello marine fish-halchery and bi-
ological station. Trans. N.Z. Inst. 38 (Art. 6 1): 1-31.
, AND T. Anderton. 1921. History of the Portobello Marine
Fish-Hatchery and Biological Station. Bull. Board Sci. Art N.Z.
No. 2.
Thomson, J. M. 1954. The Mugilidae of Australia and adjacent seas.
Aust. J. Mar. Freshwater Res. 5:70-131.
. 1966. The grey mullets. Oceanogr. Mar. Biol. Annu. Rev. 4:
301-355.
. 1975. The dentition of grey mullets. Aquaculture 5:100. (Ab-
stract.)
. 1982. The taxonomy of grey mullets, p. 1-63. /«; Aquaculture
of grey mullets. O. H. Oren (ed.). Cambridge Univ. Press, Cam-
bridge.
, AND A. E. Bennett. 1953. The oyster blenny. Omobranchus
anclius (Valenciennes) (Blenniidae). Aust. J. Mar. Freshwater Res.
4:227-233.
Thorp, C. H. 1969. A new species of mirapinniform fish (family
Kasidoridae) from the western Indian Ocean. J. Nat. Hist. 3:61-
70.
Thorsteinsson, v. 1981. The aging validation of the lumpsucker
(Cycloplerus lumpus) and the age composition of the lumpsucker
in Icelandic lumpsucker fisheries. Int. Counc. Explor. Sea CM.
I98I/G58.
Thresher, R. E., AND E. B. Brothers. In press. Reproductive ecology
and biogeography of Indo-west Pacific angelfishes. (Pisces: Poma-
canthidae). Evolution.
LITERATURE CITED
741
Tin, H. T. 1982a. Family Clupeidae, herrings, p. 64-73. In: Identi-
fication of larval fishes of the Great Lakes basin with emphasis on
the Lake Michigan drainage. N. A. Auer (ed.). Great Lakes Fish.
Comm. Spec. Publ. 82-3, Ann Arbor, MI.
. 1982b. Family Osmendae, smelts, p. 146-151. In: Identifi-
cation of larval fishes of the Great Lakes basin with emphasis on
the Lake Michigan drainage. N. A. Auer (ed.). Great Lakes Fish.
Comm. Spec. Publ. 82-3, Ann Arbor. MI.
. 1982c. Family Ictaluridae, bullhead catfishes, p. 436-457. /?!.-
Identification of larval fishes of the Great Lakes basin with emphasis
on the Lake Michigan drainage. N. A. Auer (ed.). Great Lakes Fish.
Comm. Spec. Publ. 82-3, Ann Arbor, Ml.
Tokuv.«,, K., and K, Amaoka. 1980. Studies on larval and juvenile
blennies in the coastal waters of southern Hokkaido. Bull. Fac.
Fish. Hokkaido Univ. 3 1(1): 16-49.
ToMiNAGA, Y. 1965. The internal morphology and systematic position
of Leptohrama mulleri. formerly included in the family Pempher-
ididae. Jpn. J. Ichthyol. 12:33-56.
TONG, L. J., AND R. Saito. 1977. Further postlarvae of the tarakihi
in New Zealand waters. N.Z. J, Mar. Freshwater Res. 1 1:159-162.
ToRNO, A. E., A. P. ToMO and E. Marschoff. 1977. Una nueva
especie del genero Oidiphorus McAllister y Rees (Pisces, Zoarcidae)
capturado en la adyacencia de las Islas Sandwich del Sur. Contrib.
Inst. Antarct. Argent. 223:1-10.
Tortonese, E. 1956. Plectognathi, p. 960-977. /n.Uova, larvee stadi
giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr. 38.
. 1973. Serranidae, p. 355-362. In: Checklist of the fishes of
the north-eastern Atlantic and of the Mediterranean. J. C Hureau
and Th. Monod (eds.). UNESCO, Pans.
Townsend, D. W., andJ. J. Graham. 1981. Growth and age structure
of larval Atlantic herring, Clupea harengus. in the Sheepscot River
estuary, Maine, as determined by daily growth increments in oto-
liths. Fish. Bull. U.S. 79:123-130.
ToYOSHiMA, M. 1981. Redescription of the zoarcid fish Lycozoarces
regam. Jpn. J. Ichthyol. 27:296-300.
Tranter, D.J. (ed.). 1968. Zooplankton sampling. UNESCO Monogr.
Oceanogr. Methodol. 2.
Trewavas, E. 1932. A contnbution to the classification of the fishes
of the order Apodes, based on the osteology of some rare eels. Proc.
Zool, Soc. Lond. 1932:639-659.
Troemner. J. L. 1932. Recent importations-Fundulus cubensis.
Aquarium, Philadelphia 1(3):70, 82.
. 1941. Cubanichlhys cubensis Eigenmann. Aquarium, Phila-
delphia 10(81:129-130.
Trott, L. B. 1970. Contributions to the biology of carapid fishes
(Paracanthopterygii: Gadiformes). Univ. Calif Publ. Zool. 89:1-
60.
. 1981. A general review of the pearlfishes (Pisces, Carapidae).
Bull. Mar. Sci. 31:623-629.
Trunov, I. A. 1981. Parakumba maculisquama (New genus and
species) family Macrouridae from the South Atlantic. J. Ichthyol,
(Engl, Trans. Vopr. Ikhtiol.) 21(l):27-34.
. 1982. Species of the family Regalecidae (Lampnformes) from
the Southeastern Atlantic Ocean. J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 22(1): I -6.
Tsinovskv, V. D., AND I. A. Mel'nikov. 1980. On the discovery of
Liparis koefoedi (Liparidae, Osteichthyes) in the waters of the cen-
tral Arctic Basin, p. 211-214. In: Biology of the Central Arctic
Basin. Nauka Publ., Moscow.
Tsokur, A. G. 1975. Data on the development of the Malayan Head-
light fish, Diaphus malayanus (Myctophidae, Pisces). J. Ichthyol.
(Engl. Transl. Vopr. Ikhtiol.) 15:611-622.
. 1977. Larvae of carangid fishes (Carangidae. Pisces) of the
Arabian Sea. In: Ichthyoplankton of tropical waters of the world
ocean. Acad. Sci. U.S.S.R., Proc. P. P. Shirshov Inst. Oceanol. 109:
149-227. (In Russian.)
. 1981. The \arvae of Benthosema pterota (McoQk. \%9\)Myc-
tophidae, from the Arabian Sea. J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 21(3):38-53.
Tsukahara, H. 1961. Biology of the cutlassfish, Trichiurus lepturus
Linaeus. Pan 1 . Early life history. Rec. Oceanogr. Works Jpn. Spec.
No, 5:117-121.
. 1962. Studies on habits of coastal fishes in the Amakusa
Islands. Part 2. Early life history of the rockfish. Sebastiscus mar-
moratus (Cuvier and Valenciennes). Rec. Oceanogr. Works Jpn.,
Spec. No. 6:49-55.
, and T. Shiokawa. 1957. Studies on the flying-fishes of the
Amakusa Islands. Pari 2. The life history and Yvab\l%of Parexocoetus
mento (Cuvier et Valenciennes). Sci. Bull. Fac. Agri. Kyushu Univ.
16(2):275-286. (In Japanese, English summary.)
Tucker, D.W. 1954. Report on the fishes collected by S.Y."Rosaura"
in the North and Central Atlantic, 1937-38. Part IV. Families
Carcharhinidae, Torpedinidae, Rosauridae (nov.), Salmonidae,
Alepocephalidae, Searsidae, Clupeidae. Bull. Br. Mus. (Nat. Hist.)
2(6): 163-2 14.
. 1956. Studies on the Trichiuroid fishes. 3. A preliminary-
revision of the family Trichiuridae. Bull. Brit. Mus. (Nat. Hist.)
Zool. 4(3):73-130.
Tucker, J. W., Jr. 1982. Larval development of Citharichthys cor-
nulus, C. gymnorhinus. C. spilopterus. and Elropuscrossotus(SoXh-
idae), with notes on larval occurrence. Fish. Bull. U.S. 80:35-73.
, and A. J. Chester. In press. Effects of salinity, formalin con-
centration, and buffer on quality of preservation of southern floun-
der {Paralichlhys lethostigma) larvae. Copeia.
Tung, I.-H. 1973. On the egg development and larval stages of the
grey mullet, Mugil cephalus Linnaeus. Rep. Inst. Fish. Biol. Taipei
3(1):187-210. (In Chinese, English summary.)
Turner, C H., and E. E. Ebert. 1962. The nesting of C/jrom/i ;7m«c-
tipinnis (Cooper) and a description of their eggs and larvae. Calif
Fish Game 48(4):243-248.
Turner, C. L. 1937. Reproductive cycles and superfetation in poe-
ciliid fishes. Biol. Bull. (Woods Hole) 72:145-164.
. 1940a. Superfetation in viviparous cyprinodont fishes. Copeia
1940:88-91.
. 1940b. Pseudoamnion, pseudochorion and follicular pseu-
doplacenta in poeciliid fishes. J. Morphol. 67:59-89.
. 1940c. Follicular pseudoplacenta and gut modifications in
anablepid fishes. J. Morphol. 67:91-105.
. 1940d. Adaptations for viviparity in jenynsiid fishes. J. Mor-
phol. 67:291-297.
Tyler, J. C 1968. A monograph on plectognath fishes of the super-
family Tnacanthoidea. Monogr. Acad. Nat. Sci. Phila. 16.
. 1970. Osteological aspects of interrelationships of surgeon fish
genera (Acanthuridae). Proc. Acad. Nat. Sci. Phila. 122(2):87-124.
1980. Osteology, phylogeny, and higher classification of the
fishes of the order Plectognathi (Tetradontiformes). NOA.^ Tech.
Rept. NMFS Circ. 434.
, I. Nakamura and B. B, Collette. MS. Osteology and rela-
tionships of the oceanic fish Luvanis imperialis (Luvaridae): an
acanthuroid, not a scombroid.
UcHiDA, K. 1933. Studies on the early life history of Japanese Het-
erosomata. II. Verasper variegalus. Zool. Soc. Jpn. 45:268-277.
. 1936. A note on the pelagic postlarval stages of Aniigonia
rubescens (Pisces, Capriformes). Zool. Mag. (Tokyo) 48:935-939.
, Y. DoTsu, S. MiTo AND K. Nakahara. 1958. The spawning
and early life history of "Buri" Japanese yellowtail, Seriola quin-
queradiata Temminck et Schlegel. Sci. Bull. Fac. Agric. Kyushu
Univ. 16(3):329-342.
, S. Imai, S. Mito. S. FujiTA. M. Ueno, Y. Shojima, T. Senta.
M. Tahuka and Y. Dotsu. 1958. Studies on the eggs, larvae,
and juveniles of Japanese fishes. Series I. Second Laboratory of
Fisheries Biology, Fish. Dep., Fac. Agric. Kyushu Univ.. fukuoka,
Japan. (In Japanese.)
Ueno, M., and S. Fujita. 1954. On the development of the egg of
Sillago sihama{¥orsV.a\). ix>n. J. Ichthyol. 3:1 18-120. (In Japanese,
English summary.)
.AND . 1958. Eggs, larvae, and juveniles of /'/aO'f<7'^-
alus indicus (Linne) (Platycephalidae), p. 86. In: Studies on the
742
ONTOGENY AND SYSTEMATICS OF FISHES -AHLSTROM SYMPOSIUM
eggs, larvae and juveniles of Japanese fishes. K. Uchida et al., (eds.).
Fac. Agric. Kyushu Univ., Fukuoka, Japan. (In Japanese.)
Ueno, T. 1970. Fauna Japonica: Cyclopteridae (Pisces). Academic
Press of Japan, Tokyo.
Ueyanagi, S. 1963a. Methods for identification and discrimination
of the larvae of five istiophorid species distributed m the Indo-
Pacific. Rep. Nankai Reg. Fish. Res. Lab. (I7):137-151. (In Jap-
anese, English summary.)
. 1963b. A study of the relationships of the Indo-Pacific isti-
ophorids. Rep. Nankai Reg. Fish. Res. Lab. (17):151-165. (In Jap-
anese, English summary.)
. 1 974a. On an additional diagnostic character for the identi-
fication of billfish larvae with some notes on the variations in
pigmentation, p. 73-78. In: Proc. Intematl. Billfish Symp.. Kailua-
Kona, Hawaii, 9-12 August 1972, Part 2. Review and contributed
papers. R. S. Shomura and F. Williams (eds.). NOAA Tech. Rep.
NMFS SSRF 675.
. 1974b. Present state of billfish larval taxonomy, p. 649-654.
In: The early life history of fish. J. H. S. Blaxter (ed.). Springer-
Verlag, Berlin.
, Y. NisHiKAWA AND T. Matsuoka. 1974. Artificial fertilization
and larval development of skipjack tuna, Katsuwonus pelamis. Bull.
Far Seas Fish. Res. Lab. (Shimizu) 10:179-188.
AND M. Okiyama. 1979. Morphology and phylogeny of fish
— . 1958. Studies on fish eggs and larvae of Madras Coast. Univ.
Madras, G. S. Press.
— . 1973. Studies on fish eggs and larvae from Indian waters. 1.
Development of egg and larvae of Hirundichthys (Hinmdichthys)
coromandelensis (WomeW). Indian J. Fish. 20(1):108-137.
— . 1975. Studies on fish eggs and larvae from Indian waters. 3.
larvae - scombroids. Kaiyo Kaguku (Mar. Sci.) 1 1:93-99. (In Jap-
anese.)
Ukawa, M., M. HioucHi, AND S. MiTO. 1966. Spawning habits and
early life history of a serranid fish, Epinephelus akaara (Temminck
et Schlegel). Jpn. J. Ichthyol. 13:156-161.
Umeda, S., and a. Ochiai. 1975. On the histological structure and
function of digestive organs of the fed and starved larvae of the
yellowtail, Seriola quinqueradiata. Jpn. J. Ichthyol. 21:213-219.
Utter, F. M. 1981. Biological criteria for definition of species and
district intraspecific populations of anadromous salmonids under
the U.S. Endangered Species Act of 1973. Can. J. Fish. Aquat. Sci.
38:1626-1635.
VAN der Veer, J. 1982. Simple and reliable methods for the fixation,
mounting and staining of small and delicate marine plankton for
light microscopic identification. Mar. Biol. (Berl.) 66:9-14.
Varikul, v., and C. Boonsom. 1969. Illustrations of the gross egg
and larval development of Pla Swai Pangasius sutchi Fowler. Indo-
Pac. Fish. Counc. Occas. Pap. 14:1-4.
Vatanachi, S. N. 1972. The identification offish eggs and fish larvae
obtained from the survey cruises in the South China Sea. Proc. 1 5
Session Indo-Pac. Fish. Counc, FAO, Bangkok.
Venkataramanujan, K., AND K. Ramanoorthi. 1981. Fish eggs of
Porto Novo coastal waters, east coast of India. Rapp. P. V. Reun.
Cons. Int. Explor. Mer 178:597-599.
Vergara, R. 1972. Analisis taxonomico y consideraciones filogene-
ticas sobre las especies Cubanas del genero Caranx (Teleostei, Per-
ciformes, Carangidae). Cent. Invest. Pesq. Cuba, Contrib. 34.
. 1974. Estudio fenetico de los Carangidos Cubanos. (Pisces,
Teleostei). Acad. Cienc. Cuba Ser. Oceanogr. 23:1-30.
ViALLi, M. 1937. Famiglia 2: Mugilidae, p. 433-456. In: Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
. 1956. Famiglia 3: Sphyraenidae. p. 457^61. //I. Uova, larve
e stadi giovanili di Teleostei. Fauna Flora Golfo Napoli Monogr.
38.
Victor, B. C. 1982. Daily growth increments and recruitment in two
coral-reef wrasses, Thalassoma bifasciatum and Halichoeres bi-
vittatus. Mar. Biol. (Bert.) 71:203-208.
. 1983. Recruitment and population dynamics of a coral-reef
fish. Science (Wash. D.C.) 219:419-420.
, AND E. B. Brothers. 1982. Age and growth of the fallfish
Semotilus corporalis with daily otolith increments as a method of
annulus verification. Can. J. Zool. 60:2543-2550.
VijAYARAGHAVAN, P. 1957. Studies On fish-cggs and larvae of Madras
coast. PhD thesis submitted to Univ. Madras, India.
Development of egg and early larva of Cypselurus spiloplerus (Cu-
vier and Valenciennes). Indian J. Fish. 2I(1):21 1-219.
ViLELA, M. H., AND J. J. ZijLSTRA. 1971. On the condition of herring
larvae in the central and southern North Sea. Rapp. P. V. Reun.
Cons. Int. Explor. Mer 160:137-141.
ViLLADOLiD, D. v., AND P. R. Manacop. 1934. The PhiUipine Phal-
lostethidae, a description of a new species, and a report on the
biology of Gullaphallus mirabilis Herre. Phillipp. J. Sci. 55:193-
220.
ViLLwocK, W. 1960. Uber Bastarde zwischen Angehorigen zweier
Subfamilien oviparer Zahnkarpfen (Microcyprini, Cyprinodonti-
dae). Zool. Anz. 165(9-10):349-352.
Vinagradov, K. A. 1950. Biology of the Pacific lumpfishes in Kam-
chatka waters. Priroda 39(3):69-70.
Vladykov, V. D. 1970. Pearl tubercles and certain cranial pecularities
useful in the taxonomy of coregonid genera., p. 167-193. In: Bi-
ology of coregonid fishes. C. C. Lindsey and C. S. Woods (eds.).
Univ. Manitoba Press, Winnipeg.
Vlymen, W. J. 1977. A mathematical model of the relationship be-
tween larval anchovy (Engraulis mordax) growth, prey, microdis-
tribution, and larval behavior. Environ. Biol. Fishes 2:21 1-233.
Vodyanitsky, V. a., and I. I. Kazanova. 1954. Key to the pelagic
fish eggs and larvae of the Black Sea. Tr. Vses. Nauchno-Issled.
Inst. Morsk. Rybn. Khoz. Okeanogr. 28:240-345. (In Russian.)
VON Ihering, R., J. DE Camargo B. and N. Planet. 1928. Os ovulos
e a desova dos peixes d'agua doce do Brasil. Bol. Biol, 14:97-109.
VON Westernhagen, H. 1974. Observations on the natural spawning
of Alectis indicus (Ruppell) and Caranx tgnobilis (Forsk.) (Caran-
gidae). J. Fish Biol. 6:513-516.
, AND . 1975. Rearing and spawning siganids (Pisces:
Teleostei) in a closed seawater system. Helgol. Wiss. Meeresuniers.
27:1-18.
, AND H. Rosenthal. 1976. Induced multiple spawning of reared
Siganus oramin (Schneider) (=5. canaliculatus Park). Aquaculture
7:193-196.
VooREN, C. M. 1972. Postlarvae and juveniles of the tarakihi (Tel-
eostei: Cheilodactylidae) in New Zealand. N.Z. J. Mar. and Fresh-
water Res. 6:602-618.
Voss, N. A. 1954. The postlarval development of the fishes of the
family Gempylidae from the Florida Current. I. Nesiarchus nasutus
Johnson and Gempylus Cuv. and Val. Bull. Mar. Sci. Gulf Caribb.
4:120-159.
. 1957. Fishes of the family Gempylidae collected by the Ber-
muda Oceanographic Expeditions 1929, 1930, 1931 and 1934.
Copeia 1957:304-305.
Vrat, V. 1949. Reproductive behavior and development of eggs of
the three-spined stickleback (Gasleroslens oculeatus) of California.
Copeia 1949:252-260.
Wade, C. B. 1946. New fishes in the collections of the Allan Hancock
Foundation. Allan Hancock Found. Pac. Exped. 9(8):2 15-237.
Wade, R. A. 1962. The biology of the tarpon, Megalops atlanticus,
and the ox-eye, Megalops cyprmoides. with emphasis on larval
development. Bull. Mar. Sci. Gulf Caribb. 12:545-622.
Wake, M. H. 1979. Hyman's comparative anatomy. 3rd ed. Univ.
Chicago Press, Chicago.
Wakjya, Y., AND N. Takahashi. 1913. Study on fishes of the family
Salangidae. J. Coll. Agric. Univ. Tokyo 14:265-295.
Waldron, K. D. 1968. Early larvae of the canary rockfish, Se/ia.s/o^/t'.y
pinniger. J. Fish. Res. Board Can. 25:801-803.
, AND B. M. ViNTER. 1978. Ichthyoplankton of the eastern Ber-
ing Sea. Unpubl. Rep., Northwest Alaska Fish. Cent., Seattle.
Walker, B. W. 1952. A guide to the grunion. CaHf Fish Game 38:
409-420.
LITERATURE CITED
743
Wallace, J., and R. van der Elst. 1973. Identification of juvenile
mullet. Bull. S. Afr. Assoc. Mar. Biol. Res. 10:26-27.
Wallus, R. 1981. Early development of the mooneye, Hiodon tergisus
LeSueur, with discussion of larval hiodontid taxonomy. Rapp. P.V.
Reun. Cons. Int. Explor. Mer 178:583-584.
, AND K. L. Grannemann. 1979. Spawning behavior and early
development of the banded sculpin Cottus carolinae (Gill), 1 99-
235. In: Proceedings of a workshop on freshwater larval fishes. R.
Wallus and C. W. Voigtlander (eds.). Tennessee Valley Authority,
Norris, TN.
Walters, V. 1960. Synopsis of the lampriform suborder Veliferoidei.
Copeia 3:245-247.
. 1961. A contribution to the biology of the Giganturidae with
the description of a new genus and species. Bull. Mus. Comp. Zool.
125(10):297-319.
. 1963. The trachipterid integument and an hypothesis on its
hydrodynamic function. Copeia 1963:260-270.
. 1964. Order Giganturoidei, p. 566-577. In: Fishes of the
western North Atlantic. Mem. Sears Found. Mar. Res. No. l(Pt.
4):566-577.
, AND J. E. Fitch. 1960. The families and genera of the Lam-
pridiform (Allotriognath) suborder Trachipteroidei. Calif Fish Game
46:441-451.
, AND . 1964. The ribbonfishes (family Trachipteridae)
of the eastern Pacific Ocean with a description of a new species.
Calif Fish Game 50:228-240.
Wang, J. C. 1981. Taxonomy of the early life stages of fishes — fishes
of the Sacramento-San Joaquin Estuary and Moss Landing Harbor-
Elkhom Slough, California. Ecological Analysts, Inc., Concord, CA.
Wang, J. C. S., AND R. J. Kernehan. 1979. Fishes of the Delaware
estuaries, a guide to the early life histories. Ecological Analysts,
Inc., Towson, MD.
Ward, J. W. 1957. The reproduction and early development of the
sea catfish, Galeichthysfelis. in the Biloxi (Mississippi) Bay. Copeia
1957:295-298.
Washington, B. B. 1981. Identification and systematics of larvae of
Artedius, Ctinocottus. and Oligocoltus (Scorpaeniformes: Cottidae).
Unpubl. MS thesis. Oregon State Univ., Corvallis.
, AND S. L. Richardson. MS. Systematic relationships among
cottid fishes based on developmental osteology with notes on their
allies. Gulf Coast Res. Lab., Ocean Springs, MS.
Wassersuo, R. J., AND R. K. Johnson. 1976. A remarkable pyloric
caecum in the evermannellid genus Coccorella with notes on gut
structure and function in the alepisauroid fishes (Pisces: Mycto-
phiformes). J. Zool. (Lond.) 179:273-289.
Watanabe, H., M. Yukinawa, S. Nafcazawa and S. Uevanagi. 1 966.
On the larvae probably referable to slender tuna. Allot hunnusfallai
Servently. Rep. Nankai Reg. Fish. Res. Lab. 23:85-94. (In Japanese,
English summary.)
Watanabe, M. 1960. Fauna Japonica Cottidae. Biogeographical So-
ciety of Japan, Tokyo.
. 1976. Studies on the sculpin of Japan and its adjacent waters.
Bull. Watanabe Ichthyol. Inst. No. 1 .
Watling, H., AND C. J. D. Brown. 1955. The embryological devel-
opment of the American grayling (7"Avma//i« signifer tricolor) from
fertilization to hatching. Trans. Am. Microsc. Soc. 74:85-93.
Watson, M. E., and F. J. Mather, III. MS. Species identification of
juvenile tunas (genus Thunnus) from the Straits of Messina, north-
western Atlantic and the Gulf of Mexico. Manuscript distributed
for information in connection with the informal Pacific Tuna Bi-
ology Conference, Honolulu, August 1961.
Watson, W. 1974. Diel changes in the vertical distributions of some
common fish larvae in southern Kaneohe Bay, Oahu. Hawaii. Un-
publ. MS thesis. Univ. Hawaii, Honolulu.
. 1982. Development of eggs and larvae of the white croaker,
Genyonemus linealus Ayres (Pisces; Sciaenidae) off the southern
California coast. Fish. Bull. U.S. 80:403-417.
, AND J. M. Leis. 1974. Ichthyoplankton of Kaneohe Bay, Ha-
waii: a one-year study of the fish eggs and larvae. Univ. Hawaii
Sea Grant College Program, Honolulu, Tech. Rep. UNIHI-SEA-
GRANT-TR-75-01.
Weber, M. 1913. Die Fische. Siboga Exped. 57:1-710.
, AND L. F. DE Beaufort. 1922. The fishes of the Indo-Aus-
tralian archipelago. Vol. 4. E. J. Brill, Leiden.
, AND . 1936. The fishes of the Indo-Australian Archi-
pelago. Vol. 7. E. J. Brill, Leiden.
Weihs, D., AND H. G. Moser. 1981. Stalked eyes as an adaptation
towards more efficient foraging in marine fish larvae. Bull. Mar.
Sci. 31:31-36.
Weiler, W. 1968. Fossilium Catalogus I: Animalia, Otolithi piscium
117.
Weisel, G. F. 1967. Early ossification in the skeleton of the sucker
(Catostomus macrocheilus) and the guppy (Poecilia reticulata). J.
Morphol. 121:1-18.
Weiss, G. 1974. Hallazgo y descripcion de larvas de las polaca Mi-
cromesistius australis en aguas del sector Patagonico Argentino
(Pisces, Gadidae) (Finding and description of larvae of Microme-
sistius australis in Patagonic waters of Argentina (Pisces, Gadidae).
Physis Secc. A Oceanos. Org. 33(87):537-542.
. 1975. Hallazgo, descripcion y distribucion de las postlarvas
del bacalao criollo, Salilota australis y del pez sable Lepidopus
caudatusen aguas de la plataforma argentina. Physis Secc. A Ocean-
os Org. 34(89):3 19-325.
Weitzman, S. H. 1967. The origin of the stomiatoid fishes with com-
ments on the classification of salmoniform fishes. Copeia 1967:
507-540.
. 1974. Osteology and evolutionary relationships of the Ster-
noptychidae, with a new classification of stomiatoid families. Bull.
Am. Mus. Nat. Hist. 153(3):327-478.
, AND W. L. Fink. 1983. Relationships of the neon tetras, a
group of South American freshwater fishes (Teleostei, Characidae),
with comments on the phylogeny of new world characiforms. Bull.
Mus. Comp. Zool. 150:339-395.
Wellings, S. R., and G. A. Brown. 1969. Larval skin of the flathead
sole, Hippoglossoides elassodon. Z. Zellforsch. 100:167-179.
Welsh, W. W., AND C. M. Breder. 1921. A contribution to the life
history of the puffer, Spheroides maculatus (Schneider). Zoologica
(N.Y.) 2:261-276.
Wenner, C. a. 1978. Making a living on the continental slope and
in the deep-sea: life history of some dominant fishes of the Norfolk
Canyon area. Unpubl. PhD dissertation. Coll. William and Mary,
Wilhamsburg, VA.
. 1979. Notes on fishes of the genus Paraliparis (Cyclopteridae)
on the middle Atlantic continental slope. Copeia 1979:145-146.
Werner, R. C, and J. H. S. Blaxter. 1 980. Growth and survival of
larval herring (Clupea harengus) in relation to prey density. Can.
J. Fish. Aquat. Sci. 37:1063-1069.
Westrheim, S. J. 1975. Reproduction, maturation, and identification
of larvae of some Sebastes (Scorpaenidae) species in the northeast
Pacific Ocean. J. Fish. Res. Board Can. 32:2399-241 1.
Wheeler, A. 1969. The fishes of the British Isles and north-west
Europe. Macmillan, London.
, AND J. Dunne. 1975. Triplerygion atlanticus sp. nov. (Tel-
eostei-Tripterygiidae) the first record of a tripterygiid fish in north-
western Europe. J. Fish Biol. 7:639-649.
, J. QuERO AND G. K.REFFT. 1973. Zcomorphi (Zeiformes), p.
349-354. In: Check list of the fishes of the north-eastern Atlantic
and of the Mediterranean (Clofnam). Vol. 1. J. C. Hureau and Th.
Monod (ed.). Paris, UNESCO.
White, D. S. 1977. Early development and pattern of scale formation
in the spotted sucker, Minytrema melanops (Catostomidae). Copeia
1977:400-403.
White, H. C. 1939. The nesting and embryo of Zoarcra a/igw/Z/am.
J. Fish. Res. Board Can. 4:337-338.
White, M. W., A. W. North, E. L. Twelves and S. Jones. 1982.
Early development of Notothenia neglecta from the Scotia Sea,
Antarctica. Cybium, 3rd Ser., 6(1);43-51.
Whitehead, P. J. P. 1 964. A redescription of the holotype of Clupalosa
744
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
bulan Bleeker, and notes on the genera Herklotsichthys. Sardinella
and Escuatosa (Pisces: Clupeidae). Ann. Mag. Nat. Hist., Ser. 13,
7:33-47.
. 1965. A new genus and subgenus of clupeid fishes and notes
on the genera Clupea, Spratrus and Clupeonella. Ann. Mag. Nat.
Hist. Ser. 13, 7(78):32 1-330.
. 1972. A synopsis of the clupeoid fishes of India. J. Mar. Biol.
Assoc. India 14:160-256.
. 1973. The clupeoid fishes of the Guianas. Bull. Br. Mus. (Nat.
Hist.)Zool. Suppl. 5:1-227.
. 1981. Clupeidae. In: FAO species identification sheets for
fishery purposes. Eastern Central Atlantic, Vol. 2; Fishing Areas
34, 47 (in part). W. Fischer, G. Bianchi and W. B. Scott, (eds.).
Canada Funds-in-Trust. Ottawa, Department of Fisheries and
Oceans Canada, by arrangement with FAO.
Whitley, G. P. 1 926. The biology of north-west Islet, Capricorn Group.
(C.) Fishes. Aust. Zool. 4:227-236.
. 1941. Ichthyological notes and illustrations. Aust. Zool. 10(1):
1-50.
. 1945. New sharks and fishes from Western Australia. Part 2.
Aust. Zool. 11:43^5.
. 1948. Studies in ichthyology. 13. Rec. Aust. Mus. 22(1):70-
94.
, andA, N.CoLEFAX. 1938. Fishes from Nauru, Gilbert Islands,
Oceania. Proc. Linn. Soc. N.S.W. 43(3-4):282-304.
WiCKJ-ER, W. 1957. Die Larve von Blennius fluviatilis (Asso 1784).
Biol. Zentralbl. 76:453-466. (In German.)
. 1959. Weitere utersuchungen iiber Haftfaden an Teleosteer-
eiem, speziell an Cyprinodon variegalus Lacepede 1 803. Zool. Anz.
163:90-107.
Wiley, E. O. 1981, Phylogenetics, the theory and practice of phylo-
genetic systematics. John Wiley & Sons, New York.
Williams, F. 1959. The barracudas (genus Sp/nrae^a) in British East
African waters. Ann. Mag. Nat. Hist., Ser. 2, 13:92-128,
Williams, G. C, S. W. Richards AND E. G. Farnsworth. 1964. Eggs
of Ammodytes hexapterus from Long Island, New York. Copeia
1964:242-243.
Williams, J. T. 1984. Synopsis of the pearlfish subfamily Pyramo-
dontinae (Pisces: Carapidae). Bull. Mar. Sci. 33:846-854.
, and R. L. Shipp. 1982. A new species of the genus £c/i/o</on
(Pisces: Carapidae) from the eastern Gulf of Mexico. Copeia 1982:
845-851.
Wilson, D. E. 1973. Revision of the cottid genus Gymnocanthus. with
a description of their osteology. Unpubl. MS thesis. Univ. British
Columbia.
Wilson, D. F, 1972. Diel migration of sound scatterers into, and out
of, the Cariaco Trench anoxic water, J. Mar. Res. 30:168-176.
Wilson, D. P. 1956. Nesting of the black sea-bream. Neptune (April):
15, 17.
. 1958, Notes from the Plymouth aquarium. III. J. Mar. Biol.
Assoc. U.K. 37:299-307.
. 1978. Territorial behavior of male dragonets (Catlionymns
lyra). J. Mar. Biol. Assoc. U.K. 58:731-734.
Wilson, H. V. 1891. The embryology of the sea bass (Serraniua/rar-
ius). Bull. U.S. Fish. Comm. 9(1889):209-277.
Wilson, K. H., AND P. A. Larkin. 1980. Daily growth increments in
the otoliths of juvenile sockeye salmon (Oncorhynchus nerka). Can.
J. Fish. Aquat. Sei. 37:1495-1498.
Wilson, M. V. H., and P. Veilleux. 1982. Comparative osteology
and relationships of the Umbridae (Pisces: Salmoniformes). Zool.
J. Linn. Soc. 76:321-352.
Winn, H. E., and J. E. Bardach. 1960. Some aspects of the com-
parative biology of parrot fishes at Bermuda. Zoologica (N.Y.) 45(1 ):
29-34.
Winterbottom, R. 1974a. The familial phylogeny of the Tetraodon-
tiformes (Acanthopterygii: Pisces) as evidenced by their compar-
ative myology. Smithson. Contrib. Zool. 155.
. 1974b. A descriptive synonymy of the striated muscles of the
Teleostei. Proc. Acad. Nat. Sci. Phila. 125:255-317.
, and J. C. Tyler. 1983. Phylogenetic relationships of aracanin
genera of boxfishes (Ostraciidae, Tetraodontiformes). Copeia 1983:
902-917.
Wirtz, p. 1978. The behavior of the Mediterranean Tripterygion species
(Pisces, Blennioidei). Z. Tierpsychol. 48:142-174.
WisNER, R. L. 1963. A new genus and species of myctophid fish from
the south-central Pacific Ocean, with notes on related genera and
the designation of a new tribe, Electronini. Copeia 1963:24-28.
. 1976. The taxonomy and distribution of lantemfishes (family
Myctophidae) in the eastern Pacific Ocean. Navy Oceanographic
Research Development Activity, Washington D,C,
Wollam, M. B. 1969. Larval wahoo, Acanthocybium solandeh (Cu-
vier), (Scombridae) from the Straits of Yucatan and Florida. Ba.
Dep. Nat. Resour. Leafl. 4 (Pt. 1, No. 12): 1-7.
WoNGRATANA, T. 1980. Systematics of clupeoid fishes of the Indo-
Pacific region. Unpubl. PhD dissertation. Univ. London.
. 1983. Diagnoses of 24 new species and proposal of a new
name for a species of Indo-Pacific clupeoid fishes. Jpn. J. Ichthyol.
29:385-407.
Woodland, D. J. 1983. Zoogeography of the Siganidae (Pisces); an
interpretation of distribution and richness patterns. Bull. Mar. Sci.
33:713-717.
Woods, L., and P. Sonoda. 1973. Order Berycomorphi (Beryci-
formes), p. 263-396. In: Fishes of the western North Atlantic. D.
M. Cohen (ed.). Mem. Sears Found. Mar. Res. No. 1 (Pt. 6).
WouRMs, J. P. 1972a. Developmental biology of annual fishes. I.
Stages in the normal development of Austrofundulus myersi Dahl.
J. Exp. Zool. 182:143-168.
. 1972b. Developmental biology of annual fishes. II. Naturally
occurring dispersion and reaggregation of blastomeres during the
development of annual fish eggs. J. Exp. Zool. 182:169-200.
. 1972c. Developmental biology of annual fishes. III. Pre-em-
bryonic and embryonic diapause of variable duration in the eggs
of annual fishes. J.' Exp. Zool. 182:389-414.
. 1981. Viviparity: the maternal-fetal relationship in fishes. Am.
Zool. 21:473-515.
, andO. Bayne. 1973. Development of the viviparous brotulid
fish Dinematichthys ilucoeteoides. Copeia 1973:32-40.
, and D. M. Cohen. 1975. Trophotaeniae, embryonic adap-
tations, in the viviparous ophidioid fish, Oligopus longhursli: a
study of museum specimens. J. Morphol. 147:385-402.
, and D.Evans. 1974. The embryonic development of the black
prickleback, Xiphister alropurpureus. a Pacific coast blennioid fish.
Can. J. Zool. 52:879-887.
, AND G. S. Whitt. 1981. Future directions of research on the
developmental biology of fishes. Am. Zool. 21:597-604.
Wu, H.-W. 1932. Contribution a Tetude morphologique, biologique
et systematique des poissons heterosomes (Pisces Heterosomata)
de la Chine. Thesis, Univ. Paris.
WvANSKj, D. S., AND T. E. Targett. 1981. Feeding biology of fishes
of the endemic Antarctic family Harpagiferidae. Copeia 1981 :686-
693.
Wyman,J. 1859. On some unusual modes of gestation. Am. J. Science
and Arts 27:5-13.
Yabe, H., S. Ueyanagi, S. KiicAWA AND H. Watanabe. 1959. Study
of the life-history of the sword-fish Xiphias gladins Linnaeus. Rep.
Nankai Reg. Fish. Res. Lab. 10:107-150. (In Japanese, English
summary.)
Yabe, M. 1981. Osteological review of the family Icelidae Berg 1 940,
(Pisces; Scorpaeniformes), with comment on the validity of this
family. Bull. Fac. Fish. Hokkaido Univ. 32(4):293-315.
, D. M. Cohen, K. Wakabayashi AND T. IwAMOTO. 1981. Fish-
es new to the eastern Bering Sea. Fish. Bull. U.S. 79:353-356.
-, K. Tsumura AND M. Katayama. 1980. Descnption ofa new
cottid fish, Icelinus japomcus from Japanese waters. Jpn. J. Ich-
thyology 27:106-110.
Yamamoto, K., K. Yamauchi and S. Kasuga. 1975a. On the de-
velopment of the Japanese eel, AnguiUa japonica. Bull. Jpn. Soc.
Sci. Fish. 41:21-28.
LITERATURE CITED
745
, AND T. MoRiOKA. 1975b. Pre-leptocephalic larvae of
the Japanese eel. Bull. Jap. Soc. Sci. Fish. 41:29-34.
Yamamoto. M., and K. Ueda. 1978. Comparative morphology of
fisholfaclory epithelium — 2. Clupeiformes. Bull. Jpn. Soc. Sci. Fish.
44:855-859.
Yanagawa, H. 1978. Embryonic development of the kyuriuo, Os-
merus eperlanus mordax (Mitchill). Bull. Fac. Fish. Hokkaido Univ.
29(3):195-198.
Yang, W. T., and U. B. Kjm. 1962. A preliminary report on the
artificial culture of grey mullet in Korea. Proc. Indo-Pac. Fish Counc.
9:62-70.
Yanulov, K. p. 1962. On the reproduction of the roughhead grena-
dier. Zool. Zh. 41:1259-1262.
Yapchionges, J. V. 1949. Hypomesus pretiosus ix^ <ie\e\opmtnX and
early life history. Nat. Appl. Sci. Bull. 9.
Yarberry, E. L. 1965. Osteology of the zoarcid fish Melanostigma
pammelas. Copeia 1965:442-462.
Yashouv, A., AND E. Berner-Samsonov. 1970. Contnbution to the
knowledge of eggs and early larval stages of mullets (Mugilidae)
along the Israeli coast. Bamidgeh 22:72-89.
Yasutak£, H., G. NisHi and K. Mori. 1973. Artificial fertilization
and rearing of bigeye tuna ( Thunnus ohesus) on board, with mor-
phological observations on embryonic through to early post-larval
stage. Bull. Far Seas Fish. Res. Lab. (Shimizu) (8):71-78.
Yefremenko, V. N. 1976. Morphological characteristics of Proto-
myclophum anderssoni larvae (Myctophidae). J. Ichthyol. (Engl.
Transl. Vopr. Ikhtiol.) 16:958-963.
. 1977. Morphological features of the larvae of Gvm«ojcopf/wi
opisthopterm (Mynophidae, Pisces). J. Ichthyol. (Engl. Transl. Vopr.
Ikhtiol.) 17:797-800.
. 1979a. A descnption of the larvae of Balhylagus anlarcticus
(Bathylagidae). J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 1 9( 1 ): 1 56-
160.
. 1979b. Description of the larvae of six species of the family
Chaenichthyidae from the Scotia Sea. J. Ichthyol. (Engl. Transl.
Vopr. Ikhtiol.) 19(3):65-75.
. 1979c. The larvae of six species of the family Nototheniidae
from the Scotia Sea. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 19(6):
95-104.
. 1982. The finding of fla/Zu'/agwi a«/arrt;cMi (Bathylagidae)
eggs, their characteristics and distribution in the Scotia Sea. J. Ich-
thyol. (Engl. Transl. Vopr. Ikhtiol.) 22(2): 135-1 37.
. 1983. Atlasof fish larvae of the Southern Ocean. Cybium 7(2):
1-74.
Yevseyenko, S. a. 1974. New data concerning eel eggs on Georges
Bank. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 14:294-298.
, AND V. P. Serebryakov. 1974. Larvae of Diplospinus nnil-
listnatus Maul (Pisces, Gempylidae) from the norihwestem Atlan-
tic. J. Ichthyol. (Engl. Transl. Vopr. Ikhtiol.) 14:92-98.
YuSA. T. 1954. On the normal development of the fish, Theragra
chalcogramma (Pallas), Alaska pollock. Bull. Hokkaido Reg. Fish.
Res. Lab. 10:1-15. (In Japanese, English summary.)
. 1957. Eggs and larvae of flatfishes in the coastal waters of
Hokkaido. I. Embryonic development of the starry flounder Pla-
tichthys stellalus (VaWas). Bull. Hokkaido Reg. Fish. Res. Lab. 15:
1-14.
. 1958. Eggs and larvae of flatfishes in the coastal waters of
Hokkaido. III. Post-embryonic development and larvae of the flat-
fish Lepidopsella mochigarei Snyder. Bull. Hokkaido Reg. Fish.
Res. Ub. 18:1-10.
. 1 960a. Diflferences of structures of eggs and larvae between
Limanda yokohamae Gunther and Limanda schrenki Schmidt. Bull.
Mar. Biol. Stn. Asamushi 10(2):127-131.
. 1 960b. Eggs and larvae of flatfishes in the coastal waters of
Hokkaido. IV. Embryonic development of mud dab Limanda yo-
kohamae GunlhtT . Bull. Tohoku Reg. Fish. Res. Lab. 17:15-30.
. 1960c. Embryonic development and larvae of the greenling
fish, He.xagrammus olakii Jordan and Starks. Groundfish Rep. To-
hoku Reg. Fish. Res. Lab. 26:76-80. (In Japanese.)
. 1961. Eggs and larvae of flatfishes in the coastal waters of
Hokkaido. V. Embryonic development of the flatfish. Xyslrais
ghgorjewi (Herzenstein). Bull. Tohoku Reg. Fish. Res. Lab., 19:
109-118.
. 1967. Embryonic development of the Atka mackerel, F/fu-
rogrammus aronuj Jordan et Metz. Sci. Rep. Tohoku Univ. Fourih
Ser. (Biol..) 33:301-318.
, C. R. Forrester and C. Ioka. 1971. Eggs and larvae of
Limanda yokohamae {GuMher). Fish. Res. Board Can. Tech. Rep.
No. 236.
Yuschak, P. 1982. Description ofthe eggs and larvae and osteological
development of the northern searobin, Pnonotus carolinus (Pisces:
Triglidae). Unpubl. MS thesis. Univ. Connecticut, Storrs.
Zahradica, K., and S. Frank. 1976. Observations sur la reproduction
de Nothobranchius korthausae. Aquarama 10(36):16-24.
Zahuranec, B. J. 1980. Zoogeography and systematics ofthe lan-
temfishes ofthe genus Nannobrachnun (Lampanyctini: Myctoph-
idae). Unpubl. PhD dissertation. George Washington Univ.. Wash-
ington, D.C.
Zaitsev, Y. p. 1970. Marine neustonology. Israel Program for Sci-
entific Translations, Jerusalem.
Zama, A., M, AsAi and F. Yasuda. 1977. Changes with growih in
bony cranial projections and color patterns in the Japanese boarfish,
Pentaceros japonicus. Jpn. J. Ichthyol. 24:26-34.
Zehren, S. J. 1979. The comparative osteology and phylogeny of the
Beryciformes (Pisces: Teleostei). Evol. Monogr. (Univ. Chicago) 1.
Zheng, W. 1981. Comparative studies on the scales of Chinese Ca-
rangidae. Trans. Chin. Ichthyol. Soc. 1:33-45. (In Chinese, English
summary.)
Zhitenev, A. N. 1970. Ecological and morphological features of the
reproduction ofthe lumpsucker (Cycloplerus lumpus (L.). J. Ich-
thyol. (Engl. Transl. Vopr. Ikhtiol.) 10:77-83.
ZiSMAN, L. 1982. Means of identification of grey mullet fry for culture,
p. 17-63. In: Aquaculture of grey mullets. Intl. Biol. Programme
26. O. H. Oren (ed.). Cambridge Univ. Press, Cambridge.
ZuRLiNi, G., I. FERR.ARI AND A. Nassogne. 1978. Reproduction and
growth of Euterpina aculifrons (Copepoda: Harpacticoida) under
experimental conditions. Mar. Biol. (Berl.) 46:59-64.
Zviagina, O. A. 1961. Data on the propagation and development of
fish of Laptev Sea. 1 . East-Sibenan cod (Arcotogadus bonsovi, Gad-
idae). Tr. Inst. Okeanol. Akad. Nauk SSSR 43:320-327. (In Russian
Transl. by Fish. Res. Board Can., 1969, Transl. Ser. 1278.)
Zvjagina, O. a. 1965a. Malerialy orazvitii jashcherogolobykh ryb
(Pisces, Synodontidae). Tr. Inst. Okeanol. Akad. Nauk SSSR 80:
147-161.
. 1965b. On the development of Therapon theraps Cuv. et Val.
(Pisces, Theraponidae). Tr. Inst. Okeanol. Akad. Nauk. SSSR 80:
162-166. (In Russian, English summary.)
Zvyagina, O. a., and T. S. Rass. 1977. Materials on development
of tropical carangid fishes (Carangidae, Pisces). Proc. P. P. Shirshov
Inst. Oceanol. 109:211-227.
INDEX
Italicized page numbers indicate illustrations. Index includes all genera and only higher taxa that are illustrated.
Abbottina 3 1
Ablabys 440
Ablennes 336, 345, 347, 350. 352
Abramis 131
Absalom 513
Abudefduf5'\^
Abyssocottus 409-10, 443
Acanthistius 500-1
Acanthocepola 474, 490
Acanthochromis 542, 544, 547
Acanthoclinus 467
Acanthocybium 592. 600-1. 606, 609, 612, 617-8
Acanthogobius 582
Acantholiparis 429-30
Acanthopagrus 471
Acanthophthalmus 129
Acanthopsetta 643, 654, 660
Acanthosphex 440
Acanthostracion 452, 453. 454, 459
Acanthurus 548, 549
Acentrophryne 333
Achiropichthys 103
Achiropsetta 674
Achirus 23, 25, 644, 649, 657, 664, 668
Acromycter 74, 98, 101
Acropoma 464, 466
Acrytops 629-30, 633
Adelosebastes 440
Adinia 363-6
Adioryx 18. 392
Adrianichthys 345, 352
Advent or 440
Aeoliscus 403, 404
Aesopia 644, 649, 657, 672
Aetapcus 440
Aethotaxis 563
Agonomalus 408, ¥25, 428, 442, 447
Agonopsis 408, 442
Agonostomus 54 1
Agonus 408, 442
Agrostichthys 368-9, 371-2, 379
/iWw 71, 103, 105
Ahliesaums 207-8, 21 1, 272
^;<2A:a5 578
Akarotaxis 563
.4/i«/a 52, 60, 67. 62, 98-9, 126, 138
Alcichthys 442
Aldrovandia 94-5
^/ecto511, 575, 518-9, 524
.4/epe5 511, 518, 524
Alepisaurus 200, 207-8, 212, 216, 245, 259
Alertichthys 440
Alestes 133-4
Allanetta 360
^//i/75 103
Allolepis 580
Allothunnus 592, 600-1, 670, 612, 617
^/oifl 112, 118, 124-5, 129-130, 133
Alphestes 500, 507-8
/4/u;era 452, 455-6, -^57
Amanses 452. 455, -^57
Amarsipus 621. 622-3, 625-8
Ambloplites 473
Amblycirrhitus 47 1
Amblygaster 109, 124
Ammodytes 574, 575
Ammotretis 643-4, 660, 673
Amphiprion 544
Anableps 363
Anacanthus 452, -#54, 455
Anaora 637
Anarchias 65, 71
Anarhichas 568, 577
Anarrhichthys 568
/l«r/!oa25, 29. 115, 117, 775, 125
Anchoviella 116-7, 125
Ancistrus 131, 7ii
Ancylopsetta 640. 651, 682
Andriashevia 557-8
Anguilla 56, 65, 69, 72, 94, 126
Anguillidae 87
Anisotremus 49, 489
Anodontostoma 111, 7 7(5, 124
Anomalops 392
Anoplagonus 442
Anoplarchus 57 1
Anoplogaster 7, 387, i90. 392, 490
Anoplopoma 408, '/7'#, 421-2, 442, 444-5
Anotopterus 207-8, 272, 216, 245
Antennahus 15, 44, 320, 322. i2i
/lrtr;!/a5 500, 503, 505, 509
Antigonia 393, i96, 405, 460, 489-490
Antimora 266, 280
Antipodocottus 442
Anyperodon 500
/4pe/to i99, 404
Aphanius 362, 365
Aphanopus 599-601, 604
Aphredoderus 483
Apistius 440-1
Aplatophis 71, 103, 106
Apletodon 629-30, 6ii
Aploactis 440
Aploactisoma 440
Aplocheilus 364, 366
Aplochiton 150-1, 752. 153, 202-4
Apogonidae -#70
Apogonops 464
Apolecius 530
Aporops 501, 508
Aprognathodon 103
Aplehchtus 71, 103
Aptocyclus 429-3 1 , 437
746
INDEX
747
Araiophos 185-6, 188, 190, 197
Archaulus 442
Archistes 442
Arctogadus 266, 268, 280, 284, 286-7, 290, 294-5, 297
Argentina 14. 156-7. 158-9. 161, 163^, 168-9
Argyripnus 185-6, 188, 190, 797
Argyrocetlus 442
Argyropelecus 185-8, 190-1, 197
Anoinma 622-3, 624. 625-8
Ariosoma 67, 69-70, 74
Aristostomias 171-2, 175-6, 7 79, 181, 183
Arnoglossus 19. 642-3, 648, 652, 656. 657
Arrhamphus 342-3, 352-3
Artedidraco 562. 563-4
Ariediellichthys 442
Artediellina 442
Artedielliscus 442
Artediellus 442
Artedius 408, ^/9, 421, 422. 427, 442, 446-7
Ascelichthys 419. 442
Asemichthys 442
Aseraggodes 664. 668, 672
Aspasma 629-30, 633, 6i5-<5
Aspasmichthys 629-30, 633, 6i5
Aspidophoroides 426. 428, 442
Aspredo 130
/ls5«r,ger 599-601, 604
Asterorhombus 642, 652, 686
Asterropteryx 582
/l5/rafcf 5«4, 587
Astrocottus 442
Astronesthes 171-2, 174-5, 183
ASTRONESTHIDAE 772
Astroscopus 560
Ateleobrachium 269
Atheresthes 643, 654, 659, 660. 684
Atherina 355-6. 360. 533
Atherinason 360
Athennomorus 355. 357, i59, 360
Atherinops 44. 48, 355-6. 360. i67. i67
Athehnopsis 44. 48, 355-6, i57, 360
Atherinosoma 360
Atherion 355-6, 360
Atractoscton 25
Alrophacanthus 452-4
.4/rapu5 511, 518, 524
rl/«/e511, 5/5, 517-8, 520. 524
Atypichthys 469
Auchenoceros 266, 268
Aulacocephalus 500
/lM/op«s 207-8, 2//. 258
Aulorhynchus 399. 400-4
Aulostomus 401
Austroglossus 644. 657, 665
Austrolychus 580
Austrolycichthys 578
Austromenidia 355-6
/lM.v/5 592. 600-1. 606. 670, 612, 617, 619
^zy,?*?^^^ 643. 673, 684
Bagahus 129-30,
Bairdiella 24
34. 137
Batistes 452. 457
Balistapus 452
Barbatula 129
Barbourisia 382
Bar^Mi 729
BascanichthyslX, 98. 707. 103. 107
Bathophilus 171. 174-6, 779, 181, 183
Bathyagonus 442
Bathyaploactes 440
Bathycallionymus 6311
Bathyclupea 474, 479
Bathydraco 563
Bathvgadus 274-5
Bathylagus 11, 156-7, 755, 161, 762, 163-4, 765, 167. 169
Bathyleptus 199. 201
Bathylychnops 156-7. 161, 163-4, 767
Bathymicrops 207-9, 211, 272, 256, 258
Bathymyrus 70, 73
Bathypterois 207-9, 211, 272, 256, 258
Bathysauropsis 207-8, 258
Bathysaums 207-8, 272
Bathysolea 657
Bathvstethus 469
Bathytyphlops 207-9, 272, 256, 258
Batrachocottus 410, 422, 443
Bec?c-r;a 355-7, i59, 360
7Sf'//<2/or419
Be/o«f 336, 342, i4i, 347, 350, 352
Belonion 342, i45, 347, 350, 352
Belonoperca 500
Bembradium 44 1
Bembradon 44 1
Bembras 44 1
Bemhabella 245, 2-^6, 247-50, 256
Benthenchelys 103, 106
Benthocometes 3 1 2
Benthodesmns 599-601, 602, 604, 606-7
Benthosema 218-9, 221-2, 226, 227, 229, 241-3
Bero 442
Bm'x 392
Bilabria 578
Bleekeria 574
Blepsias 408, ^^25, 428, 442, 447
Bodianus 544
Bolinichthys 218, 220-2, 226, 2i5, 236, 240-3
Bonapartia 182, 185-6, 188-9. 190-1. 79i, 195. 198
Boreogadus 266, 268, 280. 284. 286-7. 290. 294-5, 297
Borichthys 563-4
Borophryne 327, i2<S, 329, 333-4
Borostomias 171, 174-5, 183
Bostockia 469-70
BOTHIDAE 641
Bothragonus 408, -^26, 428, 442
Bothrocara 578. 580, 557, 582
Bothrocarina 578
5or;;M,s 642. 652. 655, 657, 672
Bo/;a 131, 137
Brachaluteres 452, 455
Brachydanio 131
Brachygalaxias 150, 153
Brachymystax 143-4, 7-^6, 148-9
Brachyopsis 442
748
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Brachypleura 640, 645. 648, 651, 674, 678, 681-4
Brachysomophis 103, 106
Brama 474
Branchiostegus 495
Bregmaceros 40, 42, 260-3, 266, 268, 280-2, 300. 301-8, 319
Brevoortia 40, 113, 123, 125
Brinkmanella 469, 479
Brosme 260-1, 266, 268, 277, 278. 280, 282, 284, 285, 286-9,
294-5, 297-9
Brosmiculus 266, 280
Brosmophycis 309, 3 1 1
Brotula 282, 308, 311, 319
Brotulotaenia 308, 3 1 1
Brycinus 131
Buglossidium 644. 657, 660
Bythitidae 309
Caecula 103
Caesio 468
Caesioperca 500, 503
Caesioscorpis 465, 476
Cairnsichthys 360
Callanthias 464, -/dZ, 499
Callechelys 71, 100. 103, 107
Callionyinus 44, 48, 637, 6i9
Callturwhthys 637-9
Calotomus 544, 546
Campogramma 511, 518, 523-4, 527
Cantherines 452, 455
Canthidermis 452, ^55
Canthigaster 448, -^49
Capoeta 129
Caprodon 500
Capromimus 405
Caproi 393, i95, 405, 460-1
Caralophia 103, 107
Cara«^o/fi^« 511, 513, 518, 524
Cara^A: 510, 513, 522, 524, 52S-9
Carapus 19, 309, 312-4, i;6, 317-9
Careproctus 429-32, -/i-^. 437
Carinotetraodon 448
Caristius 474
Carolophia 1 1
Caspialosa 114, 125
Catesbya 100
Catostomus 53, 129
Caulolatilus 468, 495
Caulolepis 1
Caulophryne 321. 316-1, 328. 334
Centriscidae 407
Centroberyx 392
Cenlrobranchus 218-9, 221-2, 229, 234, 240-3
Centrolabrus 544
Centrolophus 622-3, 626, 628
Centrophryne 329
Centropogon 440
Centropomus 473
Centropristis 499-50 1 , 50i, 509
Centropyge 468
Cephalopholis 500, 507-8
Cephalopsetia 641, 651, 674, 682
Ceratias 327, i29
Cera/05cope/M<r 218, 220-2, 226, 2i5. 236, 240-3
Ceratostethus 361
Cerdale 587
Celengraulis 117, /2/, 125
Cetonurus 261 , 274-5
Chaenichthys 563
Chaenocephalus 563
Chaenodraco 563
Chaenogobius 582
Chaenophryne 326, iii, 333
Chaetodiptems 465, 465, 487, 489-90
Chaetodontidae 474
Chalinura 274-5
Champsocephalus 563
Champsodon 559, 560
Channomuraena 72
C/;aA!o<r24, 126, 129, 130, 133, 7i5. 139, 206
Chascanopsetta 642, 652, 656
Chaultodus 29, 30. 170, 171-2, 174, 181, 198
Chaunax 321 , 322
Cheilinus 544
Cheilopogon 337-9, 341, i42, 343, 345, 347, i49, 352, 354
Chelidonichthys 407, 419
Chelidoperca 500
Chelmon 474
Chelonodon 448
Cheroscorpaena 440-1
Cherublemma 313, 318
Chilara 3\2-4, 319
Chilatherina 360
Chilomycterus 448, 450, 463
Chilonetus 408
Chionobathyscus 563
Chionodraco 563
Chirocentrodon 1 1 4
Chirocentrus 108-9, //5. 720, 124
Chirostoma 355-6
Chirostomias 171-2, 183
Chitonotus 419. 442
Chlorophthalmus 207-9, 2//, 256, 258
Chloroscombrus 511, 574, 518, 524
Choerodon 544
Chonerhinos 448
Choridactylus 440
Chnodorus 343, 352-3
Chromis 544
Cichlasoma 542, 54i
0//a/a 266, 268, 279, 284-90, 294-5, 297-9
Cirrhilabrus 544
Cirrhimuraena 103
Cirricaecula 103
Citharichthvs 641, 643, 650. 651-4, 670, 675, 657, 682, 685
Citharoides 641, 672-3, 676, 67<S, 681-2
C/anoi 130-1, 7ii, 134
Claringer 586
Cleidopus 392
Cleisthenes 642-3, 654, 660
Clidoderma 643, 660
Clinocottus 409, 421, 422, 427, 442, 447
Clupanodon 111, 122
C/Mpra 75. 24-6, 1 10, 118, 122-4
INDEX
749
Clupeichthys 1 1 2
Clupeoides 1 1 3
Clupeonella 110, 114
Cobitis 129, 131, 137
Coccorella 250, 257-254
Coccotropsis 440
Cocotropus 439-41
Coelorhynchus 44, 269, 274-5, 279
Co/Via 114, 118. 121. 125
Colistium 643-4, 660, 672
Coloconger 70
COLOCONGRIDAE 5i
Cololabts 14. 29, iO, 336, 338, ii9, i^2, 347, 349. 350, 352,
354
Colpichthys 360
Comephorus 408, ^^76, 442
Conger 65
Congiopodus 407, -^74. 419, 440
Congothrissa 1 1 1
Congresox 94
CONGRIDAE .57, (SJ
Congrogadus 467
Conidens 629. 630-1, 6ii
Conodon 473. 489-90
Coraanus 465
Coradion 474
Coregonus 142-4, /^^i, 148
Coreoperca 469, 47i
Corica 1 1 3
Cora 544
Coryphaena 474. 489-90, -^96-7, 498
Coryphaenoides 44, 48, 267, 269, 272, 274-5
Coltapisti4s 440
Co?//«W/a410, 443
Cottiusculus 442
Cottocomephorus 4\0, 416. 443
Cottoperca 563
Cottunculus 444
Co»M5 422, 427, 442, 446-7
Craterocephalus 360
CraUnus 500-1
Crenichthys 362-4
Cromeria 138
Cromileptes 500
Crossias 442
Crossorhombus 642, 652, 655. 657
Crossostomus 578
Cry draco 563
Cryothenia 563
Cryptopsaras 'ill, 329. 333
CryslalUas 429
Crystallichthys 429
Clenochaetus 548
Ctenochirichthys 327
Ctenolabrus 544
Ctenolucius 127. 130
Ctenopharyngodon 129, 131
Ctenotrypauchen 587
Cubaitichthvs 363, 365
Cubiccps 622-3, 62-/, 625-8
Cybiosarda 592, 600-1, 606, 612, 617
Cyclopsetta 641, 650. 651-3, 657, 670, 673-5, 682-3, 685-6
Cyclopsis 429-30
Cyclopteropsis 429-30, 437
Cydopterus 429-30. 4i2. 4i6. 437
Cyclothone 182, 185-6. 188-91. 193, 795. 198
O'f'wa 69, 72, 97
Cyematidae 9i
Cygnodraco 563
Cynoglossus 640, 6-^7. 643-4, 664, 667, 669, 670, 672, 675
Cynolebias 363-4
Cynomacrurus ll'^-S
Cynoponticus 94
Cynothrissa 1 1 3
Cyprinodon 359, 362-3, i65
Cyprinus 131, 133-4
Cypselurus 338, 341, 345, 347-8, 352, 354
Cw/wi 379, 394
Dactylanthias 500
Dactyloptena 408, 441
Dactylopterus 408. '^27, 441
Dactylopus 637
Dadyanos 578. 580
Daicocus 408
Z)a//;a 140, 142, 202
Dalophis 103, 107
Dampierosa 440
Danaphos 185-6, 188, 190, 797
Z)a«;o 131
Daruma 442
i3a.rKO""-?410, 42< 428, 444
Datnioides 465, 478
Davidjordama 578
Z)aye//a 113
Decapterus 510, 512. 57-^, 518-20, 522, 524
Dendrochirus 407. 416
Dentatherina 357
Denticeps 108-9
Derepodichthys 578, 580
Derichthyidae 57
Derichthys 96
Dermatolepis 500, 507
Dermogenys 336-8, 347, 352-3. i6/
Desmodema 368-9. 371-2, 377-9
Dexistes 643
Diaphus 207-8. 218-9. 221-2, 229, 236, 241-3
Dicentrarclms 26, 469. 509
Diceratias 330
Dicologoglossa 644, 657
Dicotylichthys 450
Dinematichthys 3 1 1
Dinoperca 465, 478
£)/oc?o« 448, '/49
Diogemchlhys 218-9, 221-2, 226, 227, 229, 241-3
Diplecogaster 629, 6ii
Diplectrum 500-1, 50i, 509
Diplocrepis 629-30, 6ii
Diplodus 14
Diplogrammus 637
Diplomvstus 1 22
Diplophus 182, 185-6, 188-9. 190-1, 792, 193. 198
Diploprion 500, 510
Diplospinus 593, 599. 600-1. 603-4, 615-7
750
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Dipterygonotus 49 1
Diretmoides 379
Diretmus 379, 390. 392, 490
Dissostichus 563
Doderleinia 464, 466
Dolichopteryx 156-7, 161, 163-4, 767
Dolichosudis 207-8
Dolliodraco 563
Dolopichthys 111, 331
Doratonotus 546
Dormitator 585. 586-7
Dorosoma 111, 118, 124
Dorsopsetta 657
Dracodraco 563
Drepane 465
Dwto 500-1
Dussumieria 110, ;/«, 720, 123-4
Dysomma 101
Dysommatidae 75, 96
Ebinania 4 1 4
£c/!e/«5 102-3, 106
Echeneis 474. 497
Echinomacrurus 274-5
Echiodon 40. 42, 309. 312, 314, i/5. 317
Echiophisll, 103, 106
Echiostoma 171-2, 174-5, 181, 183^
Ectreposebastes 407. 413. 417
frfeZ/fl 469-70
Edriolychnus 333
Ehirava 1 1 3
Eigenmannia 127, 131, ;ii. 134, 137
Elagatis 5\2, 517-20, 523-4. 526-7
Elapsopis 103
Elassichthys 335, 342, 350, 352
Elassodiscus 429
Elassoma 465-6, 469
Electrona 218-9, 221-2, 225. 226, 229. 240-243
Eleginops 563
Eleginus 266, 268, 280, 284, 286-7, 290, 294-5, 297
£/wrm 582, 584, 586-7
Eleutherochir 637
Eleutheronemus 540
EUerkeldia 500
Elops 60, 61. 62, 98-99, 126
Elytrophora 6 1 7
Embassichthys 642-3, 654, 659, 660, 66'^
Embolichthys 575
Emmelichthyops 49 1
Encheliophis 312, 314, 317-9
Enchelyopus 266, 268, 279-80. 284-9, 290. 294-5, 297. 319
Enchelyurus 570
Encrasichelina 116. 123
EngraulislA, 28. 108, 114, //S, ;2;, 122, 125
Engyophrys 642, 652, 65i, 657, 672, 674, 676, 682-3, 685-6
Engyprosopon 642, 652, 655. 657
Enneapterygius 570
Enophrys 409, -^2/, 427, 442
Eocallionymus 637
Eophycis 264
Eopsetta 642-3, 654, 659. 660, 666
Ephippus 465
Epigonus 469. ^^70, 487
Epinephelus 18. 499. 500. 507-8, 509
Epmnula 593, 594, 596, 600-1, 603
Epiplatys 359, 364
Eretmophorus 266, 268, 272, 280
Ereunias 443, 445
Erilepis 442, 444
£np5 103
Erisphex 44 1
froifl 440
fra/efc 584, 586-7
Erythrocles 473
Eschmeyer 441
Escualosa 109, 124
Etox 7-^0-;, 142, 202
Ethadophis 103
Ethmalosa 113, 125
Ethmidium 113, 122
Etropus 642, 650, 651, 674-5, 685
Etrumeus 15. 28. 29, 102, 108, 110, //.S, 720, 122-4
Eucinostomus 473
Eucitharus 641, 673-4, 676, 681, 684
Eudichthys 260-1, 263, 26^^. 266, 280
Euleptorhamphus 336-7, 352
Eumecichthys 368-9. 371-2. 277-9
Eumicrotremus iO. 31. 429-30. ■^i/-2, 434, 437
Eupleurogratnmus 599-601. 604
Euryglossa 657, 664, 665
Eurymen 4AA
Eurypegasus 402
Euryslole 355-6, i57-5
Eustomias 171-2, 175-6. /79. 181. 183
Eutaeniophorus 380-1, iiS2
Euthynnus 542. 600-1. 606. 6/0, 612, 617, 619
Evermannella 207-8, 257, 252-4
Eviota 587
Evorthodus 582
Evoxymetopon 599-601, 604
Exechodontes 578
Exocoetus 338, i-^/, 345, i'#9, 352
Expedio 584. 586
Filimanus 540
Fistularia 398. 400-1, 403-4
Flagelloserranus 508
Flagellostomias 171-2, 174, /76, 181, 183
Florenciella 469
Floridichthys 365
Fluviphylax 362
Foa -^70
Fodiaior 335, 338-9, i-//, 343, i'/9. 352, 354
Foetorepus 637
Forcipiger 21. 474
Franzia 500, 503
Fw^ 448. '/'^9
Fundulopanchax 364
Fundulus 43. 53, 359, 362-5, i65-7
Furcina 442
Ga£^e//a 266. 268. 272, 279-80
Gadiculus 266. 268, 279-80. 284, 286-7, 290. 292, 294-5, 297
INDEX
751
Gadomus 267, 272. 274-5
Gadopsis 469-70, 473, 482
Gadus 15, 29, 31, 260, 266, 268, 278. 279, 281, 284, 286-7,
290, 292. 294-7
Gadusia 1 1 2
Gaidropsarus 266, 269, 279-80, 284-90, 294-5, 297-9
Galaxias 150-1, 152. 153
Galaxiella 150-1, 153
Galeoides 540, 547
Gambusia 363-4
Gargariscus 4 1 9
Gargaropteron 21
Gargilius 279
Garialiceps 7 1
Gasterochisma 591-2, 600-1, 617-8
Gasteroclupea 122
Gasterosteus 398. 399
Gastrocyathus 629-30, 6ii
Gastropsetta 642, 651. 674, 682
Gastroseyphus 629-30, 633. 6i5
Gempylus 593-4, 599, 600-1, 603-4. 615-7
Genioliparis 429
Ge«yp/erM5 309, 312, 318
Gephyroberyx 392
Gerlachea 563
Gibber ichthys 7, 15, 391-2
Gigantactis 325. i29. 333
Giganthias 500
Gigantura 199, 201
Gilbertidia 410, 424, 427, 447
GHchnstella 113, 122, 125
G(re//a 47/
Glenoglossa 103
Glossanodon 156-7, 759, 163-4, 168-9
Glossolepis 360
Glyptauchen 440
Glyptocephalus 28. 29, 642-3, 659. 660. 664
Gnathanacanthus 440
Gnathanodon 512, J/5, 518, 524
Gnathophis 66, 70
Gnathopogon 131
Gobiesox 629-31, 633, 6i5
GOBIIDAE 55i
Gobioides 585. 587
Gobiomorphus 589-91
Gobionellus 584. 587
Gobiosoma 584-5
Gonialosa 1 1 1
Gomchthys 218-9, 221-2, 229. 234, 240-3
Gomoplectrus 500, 508, 510
Gonorhynchus 138. 139
Gonostoma 182, 185-91, 794. 195, 198
GordiichthyslX. 103, 107
GracUia 500
Grahamichthys 590
Gramma 465
Grammatobothus 642, 652, 657, 683
Grammatonotus 464-5, 467
Grammatorcynus 592, 600-1, 606, 605. 612, 617-9
Grammatostomias 171-2, 183
Grammicolepis 393
Grammistes 501, 508
Grammistops 501
Grasseichthys 138
Gulaphallus 356-1 , 361
Gunnellichthys 582, 587
Gymnammodytes 574-5
Gymnapistes 440
Gymnelopsis 578
Gymnelus 578-80, 557. 582
Gymnocaesio 468
Gymnocanthus 409, 427. 442
Gymnocorymbus 129
Gymnodraco 563
Gymnosarda 592, 600-1, 606, 609. 612, 617-8
Gymnoscopelus 218, 220-2, 226, 229. 236, 241-3
Gymnothorax 72-3
Gyrinichthys 429
Hadropareia 578
Hadropogonichthys 578
Haemulon 51, 53
Halargyreus 262, 266
Halichoeres 544
Halosauropsis 94-5
Halosaurus 94-5
Hapalogenys 465, 466, 480, 485, 487, 489-90. 493
Haplophryne 326-7, i2«. 334
Harengula 110, 775. 122-4
Harpadon 206-8, 212
Harpagifer 562. 563-4
Helicolenis 406, 410, 472. 439-40
Hemanthias 500, 503, 505, 507
Hemerocoetes 557
Hemerorhinus 103
Hemtcaranx 5\1, 518, 520. 524
Hemilepidotus 409, 47 7. 425, 427, 442, 446
Hemilutjanus 465, 480
Heminodus 4 1 9
Hemirhamphus 336-9, i40. i46. 347, i57, 352
Hemirhamphodon 337, 339, 345, 352-3
Hemitripterus 409, 421, 425, 428, 442, 446-7
Herklotsichthys 109. 123
HermosiUa 469
Herpetoichthys 103
Heterandria 362
Heterenchelyidae 5i
Heteromycteris 657, 665. 687
Helerophotus 171-2
Heteropneustes 1 30
Heterostichus 564, 570
Heterothrissa 117, 125
Hexagrammos 20, 410, 474. 421-2, 443-5
Hildebrandia 101
Hildebrandichthys 1 1 7
//;75a 111, 123-4
Himantolophus 325-7, iiO
HintoniallS. 220-1, 241-3
///0£^oi7 126, 129, 133
Hippocampus 53, i95. 400, 402, 403
Hippoglossina 642, 646. 651-2, 674, 657. 682
Hippoglossoides 29, 640, 647. 642-3, 654, 659, 660
Hippoglossus 14, 642-3, 654, 659, 660, 666
Hirundichthys 338-9, i47, 343, 352, 354
752
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Histiodraco 563
Histiophryne 'il^
Histrio 321
Holanthias 500, 505. 507
Holocentrus 18, 384, i85. 387, 392
Hoplias 133^, 136-7
Hoplichthys 408, 414, 421
Hoplolatilus 468, 495
Hoplostethus 389, 392
Hoplunnis 94-95, 96
Horaichthvs 335, 339. 345, 347, 352
Howella 465, 470, 480, 483, 487, 493
Hozukius A\Q, AAQ
Hucho 143-4, ;¥6, 148-9
Hygophum 218-9, 221-2, 226, 228. 234, 241-3
Hymenocephalus 269, 274-5, 279
Hyoinacrurus 274-5
Hypentelium 129
Hyperlophus 1 12
Hyperoglyphe 611, 625-6, 628
Hyperoplus 574. 575
Hyphalophis 103
Hyphessobrycon 133-4
Hypoatherina 360
Hypomesus 1 54
Hypophthalmichthys 129, 131
Hypoplectrus 500-1, 503
Hvpoptychus 400-4
Hyporhamphns 337-9, i42, 345, 347, 352-3
Hypsagonus 425, 426, 428, 442, 447
Hypsoblennius 3 1
Hypsopsetta 29, 643, 654, 659
Icelinus 419, 442
/ct/!/j409, 479. 442, 445
Ichthyapus 7 1
Ichrhyococcus 185-9, 191, 792. 198
Ichthyupus 103
Icichthys 29, 621-3, 625-6, 628
Icosteus 28, 29, 576-7
Ictalurusl9, 130, 134. 137
Idiacanthus 11, 171-2, 175, /«0, 181, 183-4
//(5/ia 112, 117, 720, 123, 125
lluocoetes 578-80
Indostomus 404
Inermia 49 1
Inimicus 405, 407, 419, 440
Inistius 544
/;7/jop<r 207-9, 2/2, 256, 258
Irialherina 360
/50 355, 357, i59, 360
Isopsetta 15, 642-3. 654, 660
Istiblennius 570
Istiompax 604, 605
Istiophorus 600-1, 604, 605, 607-8, 614-7
Japonolaeops 642, 686
Jeboehlkia 500, 508
Jenkinsia 33, 111, 122, 123-4
Jenynsia 363
Jordanella 363, 365
Jordania 442
7i:a/(20, 555
Kamoharaia 642, 652
Kanazawaichthys 322
Kanekonia 441
Kareius 642, 654
Kasidoron 7
Katsuwonus 24, 26, 592, 600-1, 606, 670. 612, 617, 619
Kaupichthys 100
Kertomichthys 103
Kneria 138
Knightia 122
Konosirus 111, 124
Korogaster 389
Kraemaria 588
Krefflichlhys 218-9, 221, 223, 225, 229, 241-3
Krohnius 269, 279
Krusensterniella 578
A:«w6a 274-5
Kyphosus 469, 47i
Labeotropheus 542
Labracoglossa 469, 471, 49 1
Labrisomus 53
Labroides 544
Labrus 544
Lachnolaimus 544, 545
Lactophrys 458
Lactoria 44. 48, 452
Laemonema 261, 266-7, 279-80
Laeop5 642, 652, 656
Laeviscutella 112, 125
Lagocephalus 448, 449
Lamnostoma 103
Lampadena 218, 220-2, 226, 229, 236, 240-3
Lampanyctodes 218, 220-1, 223. 226, 229, 236, 241-3
Lampanvctus 218, 220-1, 223, 226, 2i7-,S. 239, 240-3, 257
Lampichthys 218, 220-1, 223, 226, 229, 236, 241-3
Lampris 368-9, 371-2, i74-5. 377-9
Lateolabrax 465, 469, 47i. 480, 485, 487, 493, 509
La/e5 469
Latimeria 55
Le/«a 131
Leiocottus 443
Leiognathidae 465
Leiostomus 40, 42
Leiuranus 103
Lepadichthys 629-30, 633, 6iJ, 636
Lepadogaster 629-30, 6ii
Leptdion 261, 266-7, 280
Lepidoblepharon 640, 676, 678, 681-2, 684
Lepidocephalus 1 3 1
Lepidocybium 593. 594-6, 600-1, 603
Lepidogalaxias 139, 202, 205
Lepidophanes 218, 220-1, 223, 226, 2i5, 236, 240-3
Lepidopsetla 15, 642-3, 654, 659
Lepidopus 599-601, 602, 606-7
Lepidorhombus 640, 643, 646
Lepidorhynchus 274-5
Lepidotrigia 407. 419
Lepophidium 311. 312-4, 318
Leptenchelys 102-3
Leptobrama 464. 477. 479
INDEX
753
Leptocephalus 62, 63. 69-70, 94, 96
Leptocottus 409, 422, All, 443, 447
Leptolucania 363-4
Leptophilypnus 589-91
Leptosynanceia 440
Lepturacanthus 599-601, 604
Letharchus 71, 103
Lethogoleos 103
Lethotremus 429-30
Lethrinus 474
Leucichthys 143, 148
Leuciscus 133-4
Leuresthes 355-6, 360, 362
Leuropharus 103
Lhotskta 345
Lichia 512, 518, 523-4, 526, 528-9
Lile 110, 124
Limanda 640, 641. 642-3, 654, 660
Limnichlhys 557
Limnothnssa 1 1 3
Linophryne 325-6, i2<S, 329, 333-4
Liocranium 440
Lioglossina 642, 651, 674, 682
Lionurus 274-5
Liopropoma 20, 42, 500, 507. 508, 510
Liopsetta 642-3, 654, 660
L/pans 429-30, ^iA 432, 433-4. 436-7
Lipariscus 429, 432
Lipogenys 94-5
Lipogramma 465
Lobianchia 218-9, 221, 223, 226, 229, 234, 241-3
Lobotes 466. 490
Lophiodes 320
Lophius 269, 320, i2/
Lophodolos 333
Lopholatilus 18. 495
Lophonectes 642, 652, 655. 657, 683
Lophotus 19. 368-9, i70, 371-2, 377, i79
Loricaria 129-30
Loricariichthys 1 30
Lorn 261, 266, 268, 280, 284, 255. 286-9, 294-5. 297-9
Lo?e//a 266, 279-80
Loweina 19. 218-9, 221, 223, 229, 234, 240-3
Lucama 359, 363-4, i65, 368
Luciogobius 582, 55i. 585-6
Lutjanus 468. 469
Lwvan«547. 550, 591
Luzonichlhys 500, 503
Lycenchelys 578, 582
Lycengraulis 117-8, 125
Lvcodapus 578-80, 582
Lycodes 578, 580, 582
Lycodichthys 578
Lycodonus 578
Lycogrammoides 578
Lyconectes 57 1
Lyconema 578
Z,yconM.s 261-3, 267, 269, 272
Lycothhssa 1 1 7
Lvcozoarces 578
Lyopsetta 642-3, 654, 659, 660. 666
Lyosphaera 448, 450
Maccullochella 469, 473
Macquaria 469
Macristiella 209
Macristium 2 1 1
Macroparalepis 207-8, 259
Macropinnal 56-S, 161, 163-4, 767. 168
Macrorhamphosodes 452-3
Macrorhamphosus 398. 399-400, 407, -^Oi, 404-5
Macrosmia 274-5
Macrostomias 171-2, 174, 183
Macrouridae 272, 276
Macrouroides 274-5
Macrounis 269, 274-5
Macrozoarces 578-80, 55/. 582
Macrurocvttus 394, i96
Macruronus 262-3, 267, 269, 272, 281
Magalespis 5\1, 518, 524
MflA/a 274-5
Makaira 600-1, 604, 605. 607-8
Malacanthus 495
Malacocephalus 269, 274-5
Malacocottus 410, -^24, 428, 444, 446-7
Malacosteidae 779
Malacosteus 171-2, 183
Malakichthys 464
Mallotus 154-5
Malvolwphis 103, 106
Manacopus 36 1
Mancopsetta 642-4, 649, 652, 670, 672-4, 678, 681, 683-4,
686-7
Manducus 185-90, 193, 195, 198
Margrethia 182, 185-6, 188-91, 79i. 195, 198
Marlyella 643
Marukawickthys 443-5
Mastums 448, 450, 463
Mataeocephalus 274-5
Maulichthys 207-8
Maurolicus 43. 44. 185-6, 188, 190, 797. 269
Maynea 578, 580
Medialuna 469, ^^77
Megalocottus 443
Megalomycter 382
Megalops 60, 67, 62, 98-99, 126
Melamphaes 386-7. 389. 392
Melanocetus 327, iiO
Melanogrammus 31, 267-8, 279, 281, 284, 286-9, 292. 294-
5, 297
Melanonus 260-3, 266, 268, 270. 280
Melanostigma 578-80, 582
Melanostomias 171-2, 174, 775. 181, 183-4
Melanotaenia 355-7, 360-1
Melapedalion 337-43, 352-3
Membras 355-6
A/e«e -#65. 479
Menidia 26, 355-7, 362
Merlangius 267-8, 284, 286-9, 292, 294-5, 297
Merluccius 25, 29, iO. 51, 260-5, 267, 269, 272, 275. 279, 281,
283, 294, 297-9
Mesobms 269, 274-5, 277
Mesocottus 443, 445
Metacottus 443
Metavelifer 368-9, 372, i76. 378-9
754
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Metelectrona 218-9, 221, 223, 225. 229, 240-3
Metynnis 1 3 1
Microcanthus 469, 473
Microchints 644, 649, 657, 660, 664, 668. 611
Microcottus 443
Microdesmus 585. 586-7
Microgadus 266-8, 281, 284, 290, 294-5, 297
Microgobius 583, 587
Micwlabrichthys 500
Microlepidium 266
Microlophichthys 327, 331
Micromesistius 267-8, 281, 283-4, 286-7, 290, 294-5, 297
Micropercops 590
Micropterus 5 1
Microspathodon 543, 544
Microstoma 156-7, 75,S, 161. 163-4, 169
Microstomus 640, fi-/;. 642-3, 654, 659, 660, 664
Microthrissa 1 1 2
Mimasea 599
Minous 407. 414. 419
Minysynchiropus 637
Mirapinna 380, i«/
Mirophallus 361
Misgiirnus 129. 131, 137
A/o/fl 448, 450, 463
Molacanthus 448, 450
Mo/va 266, 268, 280-1, 284-289, 255, 294-5, 297-9
Monacanthus 452, 455
Monocenirus 392
Monodactylus 467
Monolene (,A2. 652, 657, 672-4, 682-3
A/ora 261, 266-8, 272. 281
Moringua 64, 70, 72, 95-6
MORINGUIDAE 93
Moroco 131
Moronf' 25, 29, 49, 469, 473. 485, 489, 508, 615-7
Mugil 24-5, 29, iO. -^^ 48, 531. 532-3, 541
MURAENESOCIDAE 81
Muraenichthys 103, 105
MURAENIDAE 89
Muraenolepis 260-3, 265-6, 268, 270. 279-80, 282, 319
Mycteroperca 500, 508
Myctophum 11, 75. 218-9, 221, 223, 226, 225-9, 234, 241-3
Myoxocephalus 409, 419, 421. 427, 443, 446-7
MyrichthyslX, 73, 103
Myripristis 384-5. 392
Myrophinae 77
Myrophis 69-70, 103, yO-/, 105
Mysthophis 103
Mystus 133-4
Nalbantichthys 578
Nanichthys 335, 340, 342, 350, 352
Nannatherina 469
Nannoperca 469-70
Nansema 156-7, 755, 161, 163-4, 169
A'aio 548, 5-^9
Naucrates 498, 510, 512, 5/6, 518, 523-4, 526, 527
Nautichthys 409, -^25, 428, 443, 447
Navodon 452, 455
Nealotus 593-4, 596, 599, 600-1, 603
Neaploactis 44 1
Neat y pus 469
Nectoliparis 429, 432
Neenchelys 103, 704, 106
Nemacheilus 129, 131, 137
Nemanthias 500
Nematolosa 111, 114
Nematistius 496-8, 522-3, 526, 527, 528, 529-30
Nematonurus 274-5
Nematops 643, 684
Nemichthyidae 55
Nemichthys 69, 72, 97
Nemipteridae 471
Neoachiropsetta 674
Neobythitinae i09
Neocentropogon 440
Neoceratias 326-7, i25
Neochanna 150-1, 153
Neoconger 69, 95, 96
Neocyema 72
Neoepinnula 593-4, 596, 600-1, 603
Neolaeops 642-3, 652, 686
Neonesthes 171-2, 183
Neoodax 544
Neoopisthopterus 112, 125
Neopagetopsis 563
Neopataecus 440
Neophos 185
Neophrinichthys 444
Neoscombrops 464
Neoscopelus 207-9, 27 7, 243, 257
Neoscorpis 465, 469, 482
Neostethus 36 1
Neosynchiropus 637
Nesiarchus 593-4, 599, 600-1, 603, 615-7
Nesogalaxias 150, 151
Nettastoma 96
Nettastomatidae 69, 55
Nezumia 267, 269, 274-5
Nicholsina 544
Mp^zow 500, 507, 510
Nomeus 621, 622-3, 625
Nomorhamphus 337-8, 352-3
Normamchthys 4\0, 416. 422, 443, 445
Notacanthus 94-5
Notemtgonus 1 3 1
Notestes 440
Nothobranchius 362-4, 368
Notolepis 207-8, 277, 218, 256
Notoliparis 429, 430
Notolychnus 218-9, 221, 223, 226, 229, 234, 240-4
Notolvcodes 578
Notoscopelus 218, 220-1, 223, 226, 229, 236, 241-3
Notothenia 563-4
Notropis 29, 131
Novaculichthys 544
Novumbra 140, 142, 202
Oce//a '^26, 428,
Ocosia 440-1
Ocyanthias 500
Ocynectes 443
OrfflX 544, 546
442
INDEX
755
Odaxothrissa 1 1 2
Odontanthias 500
Odontesthes 355-7, 359
Odontognathus 114, 125
Odontolipahs 429
Odontornacnirus 269, 274-5
Odontopysix 442
Odontostomias 174, 181, 183
Odontostomops 251, 254
Odonus 452
Ogcocephalus 321, 322
Oidiphorus 578
Oligocottus 409, 422, 427, 443, 447
Oligopliles 510, 512, 5/7, 518, 522-4, 528-9
Oligopus 3 1 1
Omosudis 200, 207-8, 272, 216, 245, 259
Ompak 130-1
Oncoptems 643, 673, 684
Oncorhynchus 20. 44, 143-4, /-/d. 148-9
Oneirodes 325, 327, ii7, 333
Onuxodon 312, 314, i;6. 318
Ophichthidae 79
Ophichthinae 77
Ophichthus 65-6, 65, 69, 71, 103, 705, 106-7
Ophidian 28, 311. 312-4, 318
Ophiodon 410, -^/^ 421-5
Ophisurus 103
Ophthalmolycus 578
Opisthonema 109, //<?, 122, 124
Opisthoproctus 156-8, 161, 164, 767
Opisthopterus 113, 118, 123, 125
Opistognathus 467
Oplegnathus 467
Opostomias 171-2, 174, 776, 181, 183
Opsarichthys 133-4
Op?/vws 389, i90
Orbonymus 637
Orcynopsis 592, 600-1, 606, 612
Oreosoma 394, i96
Orthonopias 409, 421, 427, 443
Onc/a<r 335, 339, i42, 345, 347. i49, 352, 357, 368, 533
Osmerus 29, 129, 131, 133, 153, 154
Osteodiscus 429-30
Ostichlhys 392
OSTRACIIDAE iO, ¥5i
Ostracion 452, 459
Otophidium 311. 312-4, 318
Oxyjulis 544
Oxylebius 410, '#7'^. 422, 443-5
Oxyporhamphus 44. 48. 337, i-^O. 343, 345, i46, 347, 353
Pachycara 578
Pachystomias 171-2, 183
Pagetopsis 562. 563
Pagothenia 563
Pa^rws 489-90
Paleogadus 262-5
Pallasina 408, 442
Palunolepis 47 1
Pampus 622-3, 625-8
Pangasius 1 30, 1 34
Paniolabus 5 \ 2-3, 518, 524
Papuengraulis 117, 122
Paraaploactis 44 1
Parabothus 643, 652, 686
Parabramis 133-4
Parabembas 44 1
Paracallionymus 14, 44, 48, 637, 6i9
Paracentropogon 440
Paracentropristis 500
Paracetonnrus 274-5
Parachaemchthys 563
Paraclinus 570
Paraconger 70
Paracottus 410, 443
Paradiplogrammus 637
Paradiplospinus 593-4, 600-1, 603-4
Paragalaxias 150-1
Parahemi nodus 4 1 9
Parakneria 138
Parakumba 274-5
Paralabrax 499, 500-1, 509
Paralepis 201-S, 217
Paraletharchus 103
Paralichthodes 643, 673, 684
Paralichthys 24, i9, 640. 6-^7, 642-3. 646, 648, 651-2, 674,
682, 685-6
Paraliparis 429-30, 434, 437
Paranthias 500, 507, 508, 510
Parapercis 560
Paraplagusia 643, 664
Parapsettus 465
Parapterygotrigla 4 1 9
Parasalmo 143-4, 746. 148-9
Parascorpis 469
Parasilurus 1 30
Parastromateus 5\2, 518, 527, 526. 525. 530
Parasudis 206-8, 258
Parataeniophorus 380, i5/. 382-3
Paratrachichthys 392
Paraxenomvstax 94
Parexocoetus 335, 338-9, 343, i49, 350, 352, 354
Pancelinus 409, 479. 443
Panl-fl 452, 455, 457
Parkraemaria 590
Parana 512, 518, 522-4, 528
Parophidion i 77, 312-3, 318
Parophrvs 642-3, 654, 659, 660
Parvilux 218, 220-1, 224, 2i5. 236, 240-3
Pataecus 440
Patagonotothen 562, 563-4
Paxanovia 578
Pegasidae 400
Pegasus 404
Pf,gM5a 640, 647. 644, 649. 657, 660, 664
Pelagocyclus 429. 432
Pelecanichthys 643, 652, 656, 673, 683
Pe//ona 1 1 3
Pellonula 112, 125
Pelotretis 640, 647. 643-4, 660, 667. 673, 684
Peltorhamphus 640, 647, 643-4, 660, 673, 676
Pempheris 467
Pentanemus 540
Pentherichthys 327, ii7
756
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Pepnlus 622-3, 624, 625, 627-8
Perca 29
Perca 473
Percolates 469
Percichthys 469
Percilia 469
Percis 442
Percopsis 55
Percottus 582
Perissias 642, 652, 657, 672, 674, 682-6
Penstedion 407, 419
Peristrominous 44 1
Perryena 440
Petalichthys 342, 350, 352
Phaenomonas 1 \ , 103
Phallocottus 443
Phallostethus 361
Phenacostethus 361
Philypnodon 590
F/!o/;5 577
Phottchlhys 185-8, 193
Photoblepharon 392
Photocorynus 326, 333
Photonectes 171-2, 174, 181, 183-4
Photostomias 171-2, 175-6, 779, 181, 183
Phractolemus 138
Phractura 130
Phrynichthys 333
Phrynorhombus 640, 643, 6-^6, 651
Phucocoeles 578-80
F/j.vm 264, 266, 268, 279-80, 284-9, 290, 292, 294-5, 297-9
Phyllophichthus 103
Physiculus 266-8, 272, 279-80
Ptedrabuenia 578
/'/Area 501, 508
Pisodonophis 103
Plancterus 363-4
Platamchthys 1 10
P/ara.x- 465
Platichthys 44, 642-3, 654, 659, 660
Platybelone 342, i45, 350, 352
Platycephalidae ^/-^
Platycephalns 407, 421
Plecoglossus 154. 202-4
Plectranthias 500, 503, 505
Plectrogenium 439, 441
Plectroplites 469
Plectropomus 500
Plectrypops 392
Pleuragramma 563
Pleurogrammus 410, 47-^. 421-2, 443-5
Pleuronectes 25, 260, 642-3, 660
Pleuronichthys 16, 29, 44, 48, 640, 6-^7, 643, 648, 654, 659.
660, 666
Pliosteostoma 1 1 4
Poecilia 364
Poeciliopsis 364
Poecilopsetta 643, 660, 667, 684
Poecilothrissa 1 12
Pogonolycus 578
Pogonoperca 50 1
Pogonophryne 563-4
Polistonemus 540
Pollachius 261, 267-8, 279, 281, 283-4, 286-9, 292, 294-5,
297
Pollichthys 185-89, 792, 193, 198
Polyacantiwnotus 94-5
Polydaclylits 540-1
Polyipnus 185-91, 797
Poly met me 185-9
Poly mixta 379
Polynemus 540
Poly per a 429
Polyprion 465, ¥66, 482, 485, 488, 490, 494
Potnacentrus 544
Pomatomns 47 1
Ponttmis 406-7, 413. 414
Popondetta 361
Portchthys 324
Porocottus 443
Porotnttra 387. 389. 391-2
Potamalosa 113, 122-3
Potatnorrhaphis 350
Potamothrissa 1 1 2
Priacanthidae 47-/
Prtacanthus 474
Prtonedraco 563
Prionotus 28. 29, 407, ^Z-^, 419
Pristigaster 1 1 3
Procatopus 364
froco»!/i 4 10, 422, 443
Profundulm 366
Progmchthys 338-9, 349, 354
Protnethtchthys 594, 596. 599-601, 603
Promtcrops 500, 507
Pronotogrammus 500, 503, 505. 507
Prosopium 143-5, 148, 204-5
Prosoproctus A A 1
Prosphyraeria 534
Proteracanthus 465
Protomyctophiim 218-9, 221, 224, 225, 229, 241-3, 256
Protosphyraena 534
Prototroctes 151, 752. 153, 203-4
Psatntytodiscus 643, 684
P^ews 622-3, 62-/. 625. 627-8
Psenopsis 623, 627-8
Psettichthys 642-3, 654, 659, 660
Psettina 643, 652, 65i. 657
Psettodes 640, 6-^5, 649, 672-3, 675, 676. 678
Pseiidalutarts 452, '^55. 456
Pseudamia 470
Pseudamiops 470
Pseudanthtas 500
Pseudobalistes 452
Pseudoblennius 409, 443
Pseudocalliurichthys 637
Pseudocaranx 512, 518, 526
Pseudochaentchthys 563
Psendochetltnus 544
Pseudochromidae -^67
Pseudocyttus 394
Psendogramma 501, 508, 509
Pseudolabrus 544
Pseudomugil 356-7, 361
INDEX
757
Pseudomyrophis 1 \ , 103, 105-6
Pseudonezumia 274-5
Pseudopentaceros 474, 490
Pseudophrites 563
Pseudophycis 266
Pseudopleuronectes 24, 642-3, 654, 660
Pseudopristopoma 473
Pseudorhombus 642-3, 646. 648, 651-2, 674, 682-6
Pseudoscopelus 557
Pseudosynanceia 440
Pseudotylosaurus 336, 350, 352
Psilodraco 562. 563
Psychrolules 4\0, 424. 427-8, 444, 446-7
Ptarmus 441
Pteragogus 544
Pterengraulis 1 1 7
/•?ero;5 416, 405
Pterosmaris 47 1
Pterosynchiropus 637
Pterothnssus 60, 67, 62, 99
Pterygotngla 4 1 9
Ptilichthys 565, 570
Pyramodon 312, 314, i 77. 318-9
Quassiremus 103, 106
Rachycentron 474. 489-90, -/Sd, 497-8
Raconda 1 1 3
Racoviizia 563
Radiicephalus 368-9, 371-2, i76, 378-9
Radulinus 409. ^2/, 427, 443
Rainfordia 501
Ramnogaster 1 10
Ramceps 261-2, 266, 268, 279, 284-90, 294-5, 297-9
Ranzania 448. -/^P, 450, 463
Rasirelliger 592, 594, 600-1, 606, 612. 618-9
Regalecus 368-9, 371-2, i7i. 375, 377-9
Reinhardtius 642-3. 654, 659, 660
Repomucenus 637
Reporhamphus 353
Retropinna 150-1. 752, 153, 203-4
/?e.vea 593, 596, 599-601. 603
Rhadtnesthes 171-2, 183
Rhadinocentrus 36\
Rhadulinopsis 443
Rhamphocottus 409, '/77, 422, 425, 443, 446
Rhinesomus 452. ^55. 459
Rhinocephalus 262-5
Rhinogobius 584
Rhinoliparis 429-30
Rhinoprenes 465
Rhinosardina 1 10
Rhizophryne 328
Rhodichthys 429-30
Rhomboserranus 464
Rhombosolea 643-4, 660, 673, 676, 6.^0, 684
Rhyactchthys 586. 588, 590
Rhynchactis 325-7, i29. 333
Rhynchogadus 266, 268, 272
Rhynchohyalus 156-7, 163-4, 767, 168
Rhynchorhamphus 337-8, 340, 352
Richardsomchlhys 440
Ricuzenius 443
Rimicola 629-30. 633. 6i5
i?;Va 133-4
7?/vM/i« 362-3, i65-6. 368
/?oft;a 333
Roccus 469
Rondeletia 382
Ronquilus 57 1
Rosaura 7, 199, 200, 201
Rosenblattia 469
Rosenblaltichthys 246. 247-250. 256
Rudarius 452
Ruvettus 591, 593-4, 596. 600-1. 603
i?ypncw5 501. 507. 508
Saccopharyngoidea 93
Sacura -#99. 500, 507
Salangichthys 154. 155
Sa//7ora 266, 268, 280
Salmo53. 143-4. 146. 148-9
Salvelinus 143-4. 7-^6. 148-9
Samaris 643. 660, 667. 672-3
Samahscus 643, 660, 667, 657, 684
5ar^a 592^, 600-1. 606, 609, 612. 617-8
Sardina 108-9, 123-4
Sardinella 109. 118, 720, 122-4
Sardinops 18, 29, iO, 108, 110, 77«. 123^
Sarritor 442
Satyrkhthys 4 1 9
Saurenchelys 96
Saurida 207-8, 272, 258
Sauromuraenesox 94
Scalanago 65
SCARIDAE 546
Scarus 544
Scatophagus 474
Schedophilus 621. 622-3, 625-8
Schmdleria 1 1, 552, 55i, 554
Schismorhynchus 103
Schultzea 500-1
Schultzidia 103
Scomber 24, 592, 594, 600-1, 606, 605, 615-9
5cowZ)eresox 44, 48, 335, ii6, ii9, 340, 347, 350, 352, 354
Scomberoides 5\2, 516. 518, 522-4, 528-9
Scomberomorus 591-2, 594, 600-1, 606-7. 609, 612. 617-8
Scombrolabrax 591-2. 59i, 594. 599-601. 615-7
Scombrops 474. 490
Scopelarchoides 245. 2-^6, 247-250. 256
Scopelarchus 207-8, 211, 245, 246, 247-50, 256
Scopelengys 207-8, 277
Scopeloberyx 387. 389. 392
Scopelogadus 387. 392
Scopelopsis 218, 220-1, 224, 226, 229, 236, 241-3
Scopelosaurus 207-8, 272
Scophthalmus 27, 640, 647, 643, 646, 651, 676
Scorpaena 405, 407, 47i. 414, 439, 445
Scorpaenichthys 409, 477, 425, 427, 443, 446
Scorpaenodes 407, 47i. 414
Scorpis 469
Scytalichthys 103
Sebastes 13. 23, 44, 405-6. 410, 47 7-2, 438-40, 445
758
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Sebastiscus 440
Sebastolobus 20. 405, 407, 413. 416-7, 439-40, 445
Sectator 469
Selar 5\2, 514. 518-9, 522, 526
Selaroides 512, 518, 526
Selenanthias 500, 507
Selene 512, 518, 520. 522, 526
Semicossyphus 544
Semotilus 129
Seriola 497. 498, 510, 512-3, 517. 518, 520, 523-4, 528-9
Seholella 622-3, 627-8
Seriolina 518-9, 528
Serraniculus 501, 503. 509
Serranus 500-1, 50i, 509
Serrasalmus 129, ;iO, 131, 133, 134
Serrivomeridae 87
Serhvomer 69
Setipinna 117, ;;,S, 125
Sicydium 55
Sierrathrissa 108, 114
Siganus 26, 548, 549
Sigmistes 443
Silonia 137
Siniperca 465, 469, -^Zi. 484-5, 489, 495
Snydehdia 312, 314, i/ 7, 318-9
Snyderina 440-1
5o/fa 644, 649, 657, 660, 664, 665, 672, 678
Solenophallus 361
Solenostomus 400. 402. 403-4
Solivomer 207-8, 256, 258
Sorto^a 185-8
Sparisoma 544
Spectrunctulus 3 1 1
Sphagebranchus 269
Sphagemacrurus 274-5
Sphoeroides 448, 450
Sphyraena 26, 533-4, 5i5-6, 537, 5iS-9, 541
Sphyraenodus 534
Sphyraenops 469, ¥70, 487, 490
Spinicapitichthys 637
Spiniphryne 333
Spirinchus 154. 155
Sprat el loides 111, 122-4
Spratellomorpha 1 1 2
Sprattus 108, 110, 122-4
Squalidus 131, 133-4
Squalogadus 274-5
Stegastes 544
Steindachneria 261-4, 267, 269, 277, 282-3
Stelgistrum 443
Stellerina 426, 442
S/e////i>r -^70
Stemonosudis 201-8, 217. 218, 259
Stenatherina 360
Stenobrachius 218, 220-1, 224, 2i5, 236, 241-3
Stenodus 142-5, 148
Stephanoberyx 392
Stephanolepis 452, -/f^, 455-6
Stephanophyrne 327
Stereolepis 465, -^66, 484, 489-90, 495
Sternias 443
Sternoptyx 185-6, 188, 190-1, 797
Stethojulis 544
Sthenopus 44 1
Stictorhinus 1\. 103
Stigmatonotus 465
Stlegicottus 443
Stlengis 443
Stokellia 150-1, 203-4
Stolephorus 116, 7;S, 72/. 123, 125
Stolothrissa 1 1 3
Stomias 30. 31, 770, 171-2, 174, 181. 183
Stromateus 622-3, 627-8
Strongylura 336, 339, i-^0, i42. i45. i50. 352
Stylephorus 368-9, 371-2, 375, i77, 378-9
Stylophthalmus 181
5«a'(5 20, 155, 207-8, 216, 277. 256
Sufflamen 452, 454
Sundasalanx 204
Suttonia 50 1
Svetovidovia 7, 266-8, 270, 279-81
Syacium 642, 651-3, 670, 674-5, 682, 685
Symbolophorus 218-9, 221, 224, 226, 22<S, 229, 234, 241-3
Symphodus 544
Symphurus 28. 29, 643, 664, 667, 669. 670, 672
Symphysanodon 464-5, -^66, 483, 489-90, 495
Symphysodon 542
Synagrops 464, ¥66, 490
Synanceia 440
Synaphobranchidae 75
Synaptura 644, 657
Synchiropus ^Til
Synchirus 443
Syngnathus 400. 402. 403
Synodus 16, 29, 48, 207-8, 216, 256, 258
Taaningtchlhys 2\». 220-1, 224, 229. 236, 240-3
Tactostoma 10, 171-2, 174, 776. 181, 183-4
Taenwides 582, 586-7
Taeniopsetta 642, 652, 65i, 657, 672, 682-3, 685-6
Tanakius 642-3, 654, 659, 660, 66¥
Tandanus 129
Taranetzella 578
Tarletonbeania 218-9, 221, 224, 229. 234, 240-3
Tarphops 642, 6¥6, 651, 674, 682
Tarpon 98
Taurocoltus 443
Taurulus 409
Tautoga 544
Telmatherina 355-6, 361-2
Temnocora 429
Tentoriceps 599-601, 604
Tenualosa 108, 111, 124
Tephrinectes 642, 672, 674, 678, 650. 681
Tetragonurus 622-3, 62¥, 625-8
Tetraodon 448
Tetrapturus 600-1, 604, 605, 607-8
Tetraroge 440
Tetrosomus 452
Thalassenchelys 69-70
Thalassoma 544, 5¥5
Thaleichthys 154-5
Thaumatichthvs 326-7, iiO. 333
INDEX
759
Thecopterus 443
Theragra 29, 267, 269, 281, 284, 286-7, 290, 294-5, 297.
298-9
Therapon 473
Thorophos 185-8, 190-1
Thrattidion 108, 114
Thrvssa 116, 121, 125
Thunnus 592, 600-1, 606, 610, 612, 613, 615-7, 619
Thymallus 142-4, 148, 204-5
Thvriscus 443
Thyrsites 591, 593-4, 596, 599-601, 603-4
Thyrsitoides 593-4, 600-1, 603
Thyrsitops 593, 59^^, 596, 600-1, 603
Thvsanactis 171-2
Thysanopsetta 642, 672, 674, 678, 681-2, 686
Tilesina 442
r/7Mru5 62, 63. 94
Tiluropsis 62, 6i, 70, 94
Tomeurus 364
Tongaichthys 591, 593-4, 596, 600-1, 603, 612, 615-7
ror^!<i,ge«er 448, 450
Tosana 500
Tosarhombus 643, 652, 686
Trachelochismus 629, 630-1, 6ii
Trachicephalus 440
Trachichthyidae i90
Trachichthys 389, 392
Trachidermus 443
Trachinocephalus lQl-9>, 212, 216
Trachinolus 498, 513, 5/6, 518, 523-4, 526. 528-9
Trachinus 31, 559, 560
Trachipterns 30, 368-9, i70, 372, 375, 377-9
Trachonurus 274-5
Trachurus 12, 14. 23, 510, 511, 513^, 518, 520, 526
Trachycorystes 130
Trachydermis 445
Trachyrhynchus 1()1, 269, 274-5
Trachyscorpia 440
Tragulichthys 448, '#'^9
Triacanthodes 453
Triacanthus 452, '#5i, 454
Tribolodon 129
Trichiurus 600-1, 602. 604, 606-7
Tnchodon 3\, 557. 564
Trichonotus 557
Trichopsetta 642, 652, 657, 672, 674, 682-3, 685-6
Tng/a 419
Thglops 409, 419. 443
Trigonolampa 171-2
Trinectes 640, 6-//, 644, 649, 657, 66<S, 650
Tnphoturus 218, 220-1, 224, 226, 2i5, 236, 241-3
Tnplophos 182, 185-90, 193, 198
Tripterodon 465
Tripterophycis 261, 266
Tripterygion 570
Trisopterus 267, 269, 284, 286-9, 292, 294-5, 297-9
Trisotropis 500
Trypauchen 584, 555. 587. 589-90
TMftWa 622, 626-628
Tvlosurus 336, i40. i45. 347, 350, 352
t/Zua 513, 518, 526
(7w*ra 140, 7-^7. 142, 202
Uncisudis 207-8
[/ra^;a419
t/ra5p(5 513, 518, 527. 526
Urophycis 264, 266, 268, 27<S, 279, 280-1. 284-9, 290, 294
Uropterygius 72
Valencia 364
Valenciennellus 185-6, 188, 190, 797
Fano/a 500
Velambassis 467
Velifer 368-9, 375
Ke///ror 443
Ventrifossa 274-5
Verasper 642-3, 654, 659. 660
Verecundum 642, 651, 674, 682
Verilus 464
Vinciguema 28. 29, 185-6, 188, 190, 792, 193, 797
Vomeridens 563
Winteria 156, 163
PKoo£75;a 185, 188-9, 191, 792. 193, 198
Xaniolepis 50
Xantichthys 452, -^56
Xenaploactis 44 1
Xeneretmus 408, -^26. 442
Xeneniodon 336, i'^2. 347, 350, 352
Xenisthmus 590
Xenistius 473
Xenocongridae 97
Xenolepidichthys 393
Xenornv'^/aA' 94
Xenopoecilus 345, 352
Xenopthahmchthys 156, 767, 163-4
Xi/j/j/as 49, 489, 495, 591, 600-1, 604, 606. 612, 615-7, 619
Xiphophorus 363
Xyrichthys 544, 545
A>r/a5 103
A'^'^rreun'^ 642, 646. 651, 674, 682
Yarella 185-6, 188-9, 191, 792. 193
Yirrkala 103
Yonogobius 584-5
Yo~ia 402
Zanchlorynchus 440
Zanclus 548, 550
Zaniolepis 410, '^7'^. 422, 443-5
Zaprora 57 1
Zebrasoma 548
Zebrias 644, 657
Zenarchopterus 337-8, 352-3
Zenopsis 393
Zesticelus 443
Zeugopterus 640, 643, 646. 65 1
Ze«5 398
Zoarces 578-9
Zm 368-9, i77. 372, i7i. 375, 378-9
760
ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM
Photo of symposium attendees. La Jolla, California, August 17, 1983.